Supporting Materials for NCHRP Report 626 (2009)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

NCHRP Project 10-65—Volume 2: Research Report June 2008 Part IV—Conclusions and Recommendations Final Report         PART IV – CONCLUSIONS AND RECOMMENDATIONS  315

Page intentionally left blank.

NCHRP Project 10-65—Volume 2: Research Report June 2008 Part IV—Conclusions and Recommendations Final Report CHAPTER 9 CONCLUSIONS AND RECOMMENDATIONS The overall objective of NCHRP 10-65 was to identify NDT technologies that have immediate application for routine, practical QA operations to assist agency and contractor personnel in judging the quality of HMA overlays and flexible pavement construction. This chapter provides a summary of the conclusions and recommendations resulting from the study. 9.1 Conclusions 9.1.1 Unbound Materials • The GeoGauge is a self-contained NDT device that can be readily incorporated into a QA program for both control and acceptance testing, based on the fact that: o It provides an immediate measure of the resilient modulus of the in-place unbound material. o It identified those areas with anomalies at an acceptable success rate; second only to the DSPA. o It adequately ranked the relative order of increasing strength or stiffness of the unbound materials. o It provided resilient modulus values that were correlated to the dry density over a diverse range of material types. o The normalized dispersion is less than for the other NDT devices that provide an estimate of stiffness. o The training and technical requirements for the technical are no different than what is required when using a nuclear density gauge. Two disadvantages of using this device in a QA program are the need for measuring the water content and density using other methods, which is also the case for the DSPA and other modulus estimating devices, and the need to calibrate the test results to the material and site conditions under evaluation. The latter is more important and is discussed in more detail below. The GeoGauge should be calibrated to the project materials and conditions to improve on its accuracy, especially when testing fine-grained soils. This calibration issue requires that laboratory repeated load resilient modulus tests be performed on each unbound layer for judging the quality of construction. Most agencies do not routinely perform resilient modulus tests for design. Eliminating the laboratory resilient modulus tests from the calibration procedure will reduce its accuracy for confirming the design values, but not for identifying construction defects. For those agencies that do not have access to or the capability to perform resilient modulus tests, the FHWA-LTPP regression equations can be used to calculate the target resilient modulus at beginning of construction. The target 317

NCHRP Project 10-65—Volume 2: Research Report June 2008 Part IV—Conclusions and Recommendations Final Report resilient modulus should be the value used in structural design. For the MEPDG, this is the average value measured in the laboratory. • The DSPA is also a self-contained unit that was successful in many of the areas noted above for the GeoGauge. It was the device that had the highest success rate in identifying areas with different physical conditions or anomalies. An additional advantage of the DSPA is that the results can be calibrated to the specific unbound material being tested prior to construction—when the M-D relationship is measured in the laboratory. This calibration procedure allows the DSPA to be used to detect volumetric, as well as physical, changes in the materials during construction. In other words, the DSPA modulus is measured on the M-D samples prepared at different water contents and dry densities. In short, the seismic technology can be used in day-to-day operations to assist contractor and agency personnel in judging construction and materials quality by itself or in tandem with other geophysical and/or ground truth sampling programs. Two disadvantages of the DSPA are that it consistently resulted in a higher normalized dispersion measured over a diverse range of conditions and materials, and it requires more sophisticated training of technicians to correctly interpret the load pulse and responses to ensure that a suitable test has been collected by the device. • The DCP was also successful in many of the areas noted above for the GeoGauge. However, more much more time is needed to conduct the test, especially for stiff materials and layers with large aggregate. The test results were found to be more dependent on aggregate size. The normalized dispersion was also found to be much higher than for the DSPA and GeoGauge. However, the DCP does have the capability to easily test the strength of thicker layers of unbound materials. In fact, it can be used in conjunction with the GeoGauge and DSPA in adjusting the modulus values from those devices to laboratory conditions for fine-grained soils for agencies that do not have a resilient modulus testing capability in the laboratory. Use of the DCP can be considered an option in adjusting the test results for the DSPA for those agencies that have no plans to incorporate a resilient modulus testing capability within their design or materials departments. • The GPR (single antenna method) was found to have a poor success rate in identifying anomalies. In addition, it does not provide a measure of modulus or strength of the material. More importantly, using the single antenna method requires that either the density or water content be assumed and the other parameter calculated. Both vary along the project, resulting in higher variations of the property being calculated. Using an inaccurate value can lead to an incorrect finding. For example, the GPR found some of the areas tested to have the highest density, while most other NDT devices found that area to be the softest and least dense. It was successful, however, in measuring the layer thickness of the unbound materials. Two other disadvantages of this system are in the training area and the need to calibrate the dielectric values to physical properties of the in-place material. Samples need to be 318

NCHRP Project 10-65—Volume 2: Research Report June 2008 Part IV—Conclusions and Recommendations Final Report recovered and tested to determine the water contents and densities of those areas prior to using the results for QC or acceptance. This requires that control strips be used prior to construction, and these calibration factors should be checked periodically during construction. Many agencies are not requiring control strips, or the first day of construction is the control strip. Training is another issue; this system requires more sophisticated training for the operator to interpret the measurements taken with the GPR. Thus, it is not recommended for future use in testing unbound materials to determine the quality characteristics of the in-place material. However, it is recommended that research with the GPR continue because of its continuous coverage and speed of data collection. • Similar to the GPR, the EDG was found to have a poor success rate in identifying areas with anomalies. However, this device is believed to have potential to provide volumetric data on the unbound materials for use in a QA program with continued use. The density estimated from this device is definitely related to resilient modulus across a wide range of unbound materials. However, additional data are needed to make conclusive recommendations for improving on the measurements. The variability of the water contents measured with this device was found to be very low. Other agencies are beginning to use this device in their research programs. For example, Texas and Nevada have ongoing programs that could provide improvements to the equipment and procedures in the near future. As a result, it is recommended that this device and technology be evaluated in more detail and that studies be initiated to improve its accuracy. • The LWD and FWD are believed to have limited potential for QC purposes. The LWD These devices have more potential for use in acceptance programs of the final structure, and certainly in forensic areas for evaluating the interaction between the pavement layers and foundation. The following summarizes the conclusions reached on these devices. o This technology was unable to consistently identify those areas with anomalies. o The modulus values can be influenced by the underlying layers, resulting in lower or higher and more variable modulus values. o The normalized dispersion was found to be high, relative to the other NDT devices. o The relationship between modulus from this technology and dry density was poor. o Any error in thickness of the layer being tested can result in large errors and more variability that could lead to wrong decisions being made by the contractor and agency about the construction operation. 9.1.2 HMA Mixtures • The PSPA is a self-contained NDT device that can be readily incorporated into a QA program for both control and acceptance testing of HMA mixtures. As noted above for unbound materials, an advantage of this technology is that the device can be calibrated to the specific materials being tested during the mixture design stage for HMA mixtures. This calibration procedure allows the PSPA to be used to detect volumetric, as well as physical, changes in the materials during construction. In short, the PSPA can be used in day-to-day operations to assist contractor and agency personnel in judging construction 319

NCHRP Project 10-65—Volume 2: Research Report June 2008 Part IV—Conclusions and Recommendations Final Report and materials quality by itself or in tandem with other geophysical and/or ground truth sampling programs. o The PSPA is the NDT device recommended for QA applications because it adequately identified all areas, but one, with anomalies. The PSPA provides a measure of the dynamic modulus that is needed for pavement structural designs, even before adjusting the PSPA modulus for laboratory conditions. o Similar PSPA modulus values were measured at higher temperatures when corrected for temperature using a master curve in comparison to those measured in the laboratory. o An important condition that the NDT device needs to consider is the effect of time and varying moisture content on the properties of the HMA mixture near construction and how those properties will change over time. There have been various studies completed on using the PSPA to detect stripping in HMA mixtures. For example, Hammons et al. used the PSPA (in combination with GPR) to locate areas with stripping along selected interstate highways in Georgia (Hammons et al., 2005). The test results from the Part A and B studies support a similar decision. The PSPA, however, does have some limitations regarding full-scale use in QA programs. Use of the PSPA should be delayed after rolling to allow the mix to cool. Dr. Nazarian’s recommendation is to delay all testing for one day after HMA placement and compaction. If required, this time restriction is considered a disadvantage for use in QA program. As noted above, the seismic modulus was found to be correlated to the dynamic modulus at elevated temperatures using the master curve developed from laboratory dynamic modulus tests. • A measure of the mixture density or air voids will also be required in judging the acceptability of the modulus value. However, none of the NDT methods has clear advantages for use in day-to-day construction operations. The two that deserve further evaluation include the GPR, because of full coverage in a short period of time, and the non-nuclear PQI for of safety reasons (in relation to nuclear density gauges). • The non-nuclear density gauges are also recommended for QA because they can be incorporated into control programs readily. Some contractors are already using the non- nuclear density gauges in controlling the compaction operation. This technology was also used to identify anomalies at a reasonable rate and can be used to identify tender mixtures and the effects of rolling in the temperature sensitive zone. Variations in water have a definite affect on the HMA density measured with the PQI. The manufacturer’s recommendation is to measure the density immediately after compaction, prior allowing any traffic on the HMA surface. Similar to the PSPA, this type of time restriction is considered a disadvantage to the use of the PQI in a day-to-day practical QA program. This time effect was not found within the Part A test program, but the moisture effect was observed within Parts A of the field evaluation. Use of other non- nuclear density gauges (PaveTracker) did not exhibit this moisture sensitivity. However, the effect of water on these gauges was not included in the field evaluation as a primary 320

NCHRP Project 10-65—Volume 2: Research Report June 2008 Part IV—Conclusions and Recommendations Final Report variable. Measurements were taken after heavy rains in areas where the readings were previously taken prior to the thunderstorms. The same density values were measured, but after removing and drying all free water at the surface. This potential bias of free water on the surface is not considered a limitation but must be considered in taking measurements for control purposes. • Use of the GPR technology using the single antenna method, even with mixture calibration, requires assumptions on specific volumetric properties that vary along a project. Using the multi-antenna method is expected to improve on the measurement of the volumetric properties and identification of areas with deficiencies or anomalies. Thus, the GPR is suggested for continued research studies, especially with the multiple antenna system, which is a proprietary analysis system. The proprietary system needs additional validation prior to full-scale implementation into a QA program. • The FWD is not recommended for use in QA programs, because this technology was unable to identify some of the anomalies. In addition, the FWD has high variation in elastic modulus values, and those values are influenced by the strength of the underlying materials and layers. 9.2 Recommendations 1. It is recommended that a project be procured and developed to sponsor the production of continued improvements to the steady-state vibratory and ultrasonic technologies. 2. The IC or instrumented rollers can be valuable to a contractor in terms of controlling the compaction operation. These rollers that operated without problems were used on too few projects to recommend that they be immediately included in QA programs. However, their use can assist the contractor in optimizing the compaction of the material. Their disadvantage for HMA layers is the temperature of the mat issue. Decreases in temperature will cause the stiffness of the mat to increase. Thus, other devices still need to be used with the IC rollers for control. The IC rollers are not recommended at this time for acceptance. 3. Research with the multi-antenna GPR device and proprietary data interpretation system should not be abandoned and should be validated in future studies. This system definitely shows promise in providing the volumetric properties for HMA mixtures. The data can be collected at highway speeds and the proprietary data interpretation system can provide results on a real-time basis. The disadvantage of this system is that it also needs field cores for calibrating the method to project specific conditions. These cores should be taken periodically to confirm the calibration factors being used in estimating the volumetric properties. 9.3 Impediments to Implementation of Recommended Technologies Initially, the availability of the NDT equipment can be an impediment to implementing the recommended QC/QA results, although this is a relatively minor obstacle. A more 321

NCHRP Project 10-65—Volume 2: Research Report June 2008 Part IV—Conclusions and Recommendations Final Report 322 challenging impediment is adequately training personnel in the use of the equipment and data interpretation procedures. Some of the devices will be easy to implement (for example, the DCP), while others will require additional work (for example, interpretation of the GRP data). To address this challenge quickly, it will be necessary to coordinate with training agencies such as the National Highway to develop training programs to accelerate implementation of these products.