Intelligent Soil Compaction Systems (2010)

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.

0 9.1 Overview NCHRP Project 21-09, “Intelligent Soil Compaction Sys- tems,” involved extensive field testing and analysis to better understand roller-integrated CCC and IC and to develop recommended specifications for the use of roller-integrated CCC in QA of soil and aggregate base material compaction. The quality of constructed earthwork materials is clearly critical to the performance of pavements. Roller-integrated measurement of soil properties holds significant promise in that it provides an effective tool to assess the quality of earthwork construction. The full or complete coverage capa- bility of roller-integrated CCC is a significant improvement over current spot-test-based QA. The use of roller-integrated CCC enables state departments of transportation (DOTs) to enforce high expectations for earthwork quality. The success- ful implementation of roller-integrated CCC for earthwork QA will require high expectations as well as coordination and buy-in from DOTs and contractors. QA personnel and con- tractors will need proper training on the capabilities and use of roller-integrated CCC. Intelligent compaction, currently implemented using au- tomatic feedback control of vibration amplitude and some- times frequency, is in its infancy and will likely evolve to incorporate numerous ways in which the process of compac- tion is improved and made more efficient. The capabilities and user-friendliness of onboard PCs and software will also likely improve significantly. Finally, the measurement sys- tems, currently limited by the influence of operating param- eters, local heterogeneity, and measurement depths that far exceed lift thickness, will continue to evolve to account for these factors. True performance-based assessment of earthwork materi- als using roller-integrated CCC is within reach. The results presented here illustrate reasonably complex but determin- istic soil behavior within the measurement volume of a vi- brating roller. With further research, instrumented rollers will likely become capable of measuring the requisite layered earthwork properties to predict performance of pavement systems. The specific findings and conclusions from this study are summarized below in sections consistent with the presentation of the report. 9.2 Review of Literature and European CCC Specifications Roller-integrated CCC has been used in European practice since the late 1970s, whereas IC technology has been available commercially only since the late 1990s. The initial research on roller-integrated measurement dates to the 1970s in Sweden with the development of the CMV by Geodynamik. Roller MVs have evolved over the past 30 years within the roller manufacturing community. There are currently a handful of proprietary vibration-based MVs in the market. When Geo- dynamik first introduced the compaction meter and CMV, vibration was implemented mechanically via a two-piece ec- centric mass assembly (clamshell) within the drum. If rotated in one direction, the two eccentric masses would join together and provide maximum eccentric force or theoretical ampli- tude A. If operated in the reverse rotational direction, A would be a minimum. Automatic feedback control–based IC was made possible by the introduction of new drum technology in the 1990s. Bomag introduced the Variocontrol roller with counterrotating eccentric masses and servo-hydraulic control of the vertical component of eccentric force. Similarly, Am- mann introduced the ACE roller with servo-hydraulic two- piece eccentric mass moment control and frequency control. Other manufacturers have followed suit. The integration of GPS-based positioning with roller-based measurement of soil properties, and the incorporation of user-friendly por- table PCs with graphical software, has evolved over the past 5 to 10 years. c h a p t e r 9 Conclusions

  A considerable body of literature exists pertaining to roller vibration, roller-integrated measurement systems, roller modeling, and correlation of roller MVs to spot-test measurements. The modeling of vibratory roller-soil inter- action has evolved with experimental studies of roller be- havior. The initial assumptions about continuous drum/soil contact in the 1970s and 1980s gave way to experimental observation and lumped parameter modeling of partial loss of contact, chaotic drum jumping, and rocking mode in the 1990s and 2000s. Both experimental correlation studies and numerical investigations have shown that vibration-based roller MVs all track increases in soil stiffness/modulus ef- fectively with the exception of CMV and CCV below values of approximately 8 to 10. Based on limited studies found in the literature, the measurement depth of roller MVs was found to vary from 0.6 to 1.0 m deep for 10-ton vibratory rollers to 0.8 to 1.5 m for 12- to 17-ton vibratory rollers. A commonly observed rule of thumb in the compaction com- munity is that each 0.1 mm of amplitude A corresponds to 0.1 m of measurement depth. Owing to this measurement depth, a number of experimental studies have shown that roller MVs reflect the properties of both the compaction lift material and the sublift material. A number of studies have successfully correlated roller MVs to spot-test mea- surements. Given that roller vibration-based MVs reflect soil stiffness, roller MVs correlate well with PLT and LWD moduli. Successful correlations have been found between roller MVs and dry density. Specifications for QA of earthwork compaction using roller-integrated CCC were first introduced in Austria (1990), Germany (1994), and Sweden (1994). Revisions to these original specifications have been made in each country. The International Society of Soil Mechanics and Geotech- nical Engineering (ISSMGE) recently adopted the Austrian specifications for CCC. The Austrian/ISSMGE and German specifications each permit multiple options for using CCC in earthwork compaction QA. The most common and sim- plest approach (and the only approach permitted in Sweden) uses CCC to identify weak areas for evaluation via static PLT, LWD, or density spot testing. Acceptance is based on these weak areas meeting prerequisite PLT modulus, LWD modu- lus, or density requirements. The most advanced CCC-based QA specifications involve correlating roller MVs to PLT modulus, LWD modulus, or density in a defined calibration area. If a suitable correlation is found, a target roller MV is determined from the MV versus spot-test regression equa- tion. Acceptance is based on comparison of roller MV data collected in a production area to the target roller MV. Based on a survey of European practice, the calibration approach is challenging to implement and requires a high level of on-site knowledge. 9.3 Fundamentals of Roller-Based Measurement Systems A detailed investigation of numerous roller-based mea- surement systems was conducted to characterize the ground surface area represented by single-roller MVs, the spatial resolution in roller MV records, and uncertainty in roller MVs. Independent evaluation was performed to reproduce and validate numerous roller MVs using independent instru- mentation. Extensive testing was performed to characterize the influence of roller operational parameters—namely, ec- centric force amplitude, vibration frequency, roller speed, and forward versus reverse driving mode—on roller MVs. The objectives of these studies were to improve fundamental understanding of roller-based measurement systems and to provide guidance for roller-integrated CCC specifications. Each vibration-based roller MV investigated—Ammann and Case/Ammann k s , Bomag E vib , Dynapac CMV D , and Sakai CCV—provides a measure for an area of soil the width of the roller (2.1 m) by a spatial distance in the direction of travel that varies across MVs (0.06 to 1.0 m was observed). The re- porting resolution of roller MVs varied from 0.2 to 1.0 m. To ensure full coverage with roller-integrated CCC, the spatial distance over which a single-roller MV is reported should equal the reporting resolution. The reporting spatial resolu- tion should be no less than 10 times the GPS accuracy. Real- time kinematic (RTK) differential GPS (accuracy ~1 to 2 cm) is recommended for use with CCC and IC rollers. Assuming RTK accuracy, the spatial resolution of roller MVs should be no less than 0.25 m. The GPS-based position reporting of roller MVs exhibited errors of 0.4 to 1.5 m for the rollers and measurement systems investigated. This error was due to av- eraging of roller vibration data within each reported roller MV and to latency in onboard computation. A position- reporting procedure was developed and is recommended for roller-integrated CCC specifications (see Chapter 7). A series of tests were performed with all rollers and roller MVs to characterize the uncertainty with which single-roller MVs should be reported. Repeatability testing of properly working CCC/IC rollers and roller measurement systems often revealed an uncertainty of ±10% (one standard devia- tion); that is, a repeated pass over the same area will yield in- dividual roller MVs within ±10% of MVs from the previous pass. Repeatability testing plays an important role in verify- ing the proper working condition of a vibratory roller and/or roller measurement system. A repeatability testing procedure was successful in identifying when a roller or roller measure- ment system was faulty. Repeatability testing of pad foot mea- surement systems revealed single MV uncertainties of ±50% to 100%. A repeatability testing procedure was developed for CCC specifications and is further described in Chapter 7.

 Roller MV uncertainty will influence the establishment of QA criteria for spatial differences in pass-to-pass roller MVs as described in Chapter 7. Roller-integrated measurement of soil stiffness is not cur- rently standardized (i.e., magnitude and rate of loading, etc.). With CCC and IC rollers, measurement occurs during roller operation, and roller operation parameters can vary consid- erably. Field testing was performed to characterize the influ- ence of various operational parameters on roller MVs and to determine if these influences are predictable and there- fore could be accounted for. Field testing revealed that the influence of the magnitude of eccentric force (or theoretical drum vibration amplitude A) on roller MVs varies widely. From low to high A vibration on the same material, roller MVs were found to change by as much as 100%. The ampli- tude dependence of roller MVs was not determinate and not predictable. Roller MVs were found to increase, decrease, or remain the same with increasing A depending on the soil and layering conditions. The mechanics-based understanding of this is discussed in Section 9.4. Due to the unpredictability in A dependence, a fixed-vibration amplitude is recommended for roller-integrated CCC. Evaluation of travel speed dependence on roller MVs pro- duced mixed results. CCV and CMV D were found to decrease noticeably with an increase in roller speed. The influence of roller speed on E vib was inconclusive within the uncertainty of the measurement approach and the limited data collected. Roller MVs were found to be mildly dependent on forward- versus reverse-driving modes. Roller MVs differed by 2% to 13% in forward- versus reverse-driving mode. Given that typical compaction work involves forward- and reverse- driving sequences, there is considerable benefit to employ- ing roller MVs in both forward and reverse modes. Forward and reverse mode measurement can be considered; however, site-specific calibration is required to characterize and verify the relationship between forward-measuring and reverse- measuring roller MVs. The vibration-based roller MVs investigated—Ammann and Case/Ammann k s , Bomag E vib , Dynapac CMV D , and Sakai CCV—correlate well with each other over the range of soft to stiff soil conditions investigated in this study. CCV and CMV were relatively insensitive to changes in soil stiffness below values of about 8 to 10, consistent with findings in the literature (see Chapter 2). Many of the roller MVs employed by manufacturers were validated using independent instru- mentation and implementation of published roller MV al- gorithms. This dispels the “black box” mentality that would inhibit implementation within the engineering community. Local soil heterogeneity perpendicular to the direction of roller travel has a significant influence on roller MVs. Due to the nature of drum instrumentation, roller MVs are di- rectionally dependent on heterogeneous soil. Bidirectional roller MVs were found to vary by 100% due to transverse soil stiffness variability. This was confirmed by LWD test- ing across the drum lanes. As a result, spot testing should be conducted across the drum lane when correlating to roller MVs, and great care should be used when performing spatial statistical analysis of pass-to-pass data maps in the presence of heterogeneity. 9.4 Relationship Between Roller- Measured Stiffness and In-Ground Response To better characterize what roller MVs reflect and to build a framework for performance-based assessment of earth- work properties using roller-based measurements, a series of test beds were constructed with in-ground instrumentation to capture in situ stress-strain-modulus behavior. Numer- ous vertically homogeneous embankments and layered sub- grade, subbase, and base test beds were instrumented with stress and strain sensors at multiple levels to capture in situ behavior during static and vibratory roller passes. This as- pect of the study sought to explain, from a mechanics-based perspective, the vibration amplitude dependence of roller MVs, characterize measurement depth of the instrumented roller, and determine how current roller MVs are related to in situ soil response in vertically homogeneous and layered structures. Roller MVs measure to depths considerably greater than typical compaction lifts. For the vertically homogeneous em- bankment conditions and the 11- to 15-ton smooth drum vi- bratory rollers used in this study, the volume of soil reflected in a roller MV is cylindrically shaped, extending to 0.8 to 1.2 m deep and 0.2 to 0.3 m in front of and behind the drum. Drum/soil contact widths range from 0.1 to 0.3 m and de- crease as soil stiffness increases. For vertically homogeneous embankment situations, the measurement depth of roller MVs is controlled by the relative decay of roller-induced cy- clic stress and strain and is reached when values have decayed to 10% of their peak. The measurement depth was mildly influenced by vibration amplitude; that is, a 0.1-mm increase in A yielded about a 3.0-cm increase in measurement depth. This is significantly different than the 0.1-mm and 10.0- cm rule of thumb found in the literature. Roller MVs are a composite reflection of typical base, subbase, and subgrade structures with surface-to-top-of-subgrade thickness of less than approximately 1 m. The contribution of each layer to the roller MV is influenced by layer thickness, relative stiffness of the layers, vibration amplitude, and drum/soil interaction issues (contact area, dynamics). The contribution of sublift materials to roller MVs can be significant. In situ stress-strain-modulus measurements at depths to 1 m beneath the roller revealed highly nonlinear modulus be-

  havior within the bulb of soil reflected in roller MVs. In base, subbase, and subgrade structures, modulus varies widely from layer to layer and within layers. Modulus values increased by a factor of 2 with depth in vertically homogeneous embank- ment test beds. In situ modulus is strongly influenced by vi- bratory loading. A change in vibration amplitude from low to high created a twofold change in modulus. Numerical simu- lation of roller-soil interaction using modulus function pa- rameters for cohesive and cohesionless soils revealed signifi- cant variations of the modulus field beneath the roller. Plane strain conditions exist under the center of the drum and do not exist under the drum edges. As a result, the soil under the drum center responds stiffer than the soil under the edge. The nonlinear modulus behavior and three-dimensional nature of soil behavior beneath the drum must be considered in the extraction of performance-related parameters from roller- based measurements. The amplitude (A) dependence of roller MVs—particu- larly stiffness measures such as E vib and k s —is a result of stress-dependent soil modulus, layer interaction, and drum/ soil contact mechanics. For vertically homogeneous embank- ment conditions, the nature of the MV-A dependence (i.e., positive, negative, or neutral) depends on the modulus func- tion parameters of the soil. Granular soils that are governed by mean effective stress-induced hardening may generally ex- hibit a positive roller MV-A dependence (i.e., an increase in A yields an increase in roller MV). Conversely, cohesive soils governed by shear stress–induced softening may generally exhibit a negative roller MV-A dependence (i.e., an increase in A yields a decrease in roller MV). The roller MV-A de- pendence of layered structures is more complex and is influ- enced by stress-dependent soil modulus (modulus function parameters), layer thickness, relative stiffness of layers, and drum/soil interaction issues. Both positive and negative roller MV-A dependence is possible, even within the same material. The roller MV-A relationship is site dependent and cannot be predicted a priori. Roller MVs were found to be insensitive to the compaction of thin lifts [e.g., 15 cm (6 in)] of stiff base material placed directly over a soft subsurface. Roller MVs were sensitive to compaction of 30-cm lifts of the same stiff material over soft subgrade. The sensitivity of roller MVs to compaction of thin lifts improves as the modulus ratio of the overlying to under- lying layers decreases, and roller MVs were sensitive to com- paction of 15-cm lifts of base material placed atop similar base material. These results imply that CCC-based QA of thin base layers atop softer subgrade may be unreliable. Levels of vibratory roller-induced deviator stress were found to be considerably greater than those used in labora- tory M r testing, whereas levels of confining stress were con- siderably less. Even during low excitation force associated with finishing passes and proof rolling of compacted soil, es- timated deviator stresses q from z = 0 to 0.5 m (0 to 1.6 ft) in clayey sand were up to three times greater than the maximum q values used for laboratory M r testing of subgrade soils. Sim- ilarly, estimated q values in crushed-rock base course were up to 2.5 times greater than the maximum q used for laboratory M r testing of base materials. For z > 0.5 m, field and maxi- mum laboratory q values were reasonably similar. Conversely, values of p observed in the field were approximately 0.3 to 0.5 of those used during laboratory M r testing. The extraction of mechanistic material parameters using roller-based measurements for performance-based specifi- cations consistent with mechanistic-empirical-based design (e.g., AASHTO 2007 Pavement Design Guide) is possible but challenging. The extraction of appropriate parameters must account for the three-dimensional nature of the roller/soil interaction, the influence of layers, the nonlinear modulus of each involved material, and the dynamics of the drum/soil interaction. 9.5 Evaluation of Automatic Feedback Control–Based Intelligent Compaction Current technology for IC involves sensing via measure- ment of vibration-based parameters and adapting via AFC of excitation force amplitude (Ammann, Bomag, Case/Am- mann, Dynapac) and in some cases excitation frequency (Ammann, Case/Ammann). At a minimum level, each man- ufacturer controls the vertical excitation force amplitude to prevent unstable “jump” mode vibration of the roller (see Section 2.1.2). When the measurement system senses jump mode, the vertical excitation amplitude is decreased until the measurement system indicates stable vibration. Jump mode may persist even at the lowest vibration amplitude setting. This level of AFC is aimed at protecting the roller from ac- celerated wear and the operator from chaotic response of the roller. Bomag, Ammann, and Case/Ammann use additional AFC in an attempt to improve compaction and uniformity. Bomag controls the vertical excitation amplitude based on the relationship of the current roller MV to a limit MV. The vertical excitation is maximized within one of five levels cho- sen by the operator and is decreased if MV is greater than or equal to the limit MV. Ammann and Case/Ammann control the excitation amplitude and frequency to maintain one of three levels of force transmitted to the soil. The principles of AFC-based IC were investigated using the Bomag Variocontrol and Ammann ACE systems. Repeated passes were performed with both IC rollers over mixed mate- rial test beds exhibiting significant longitudinal variability in ground stiffness (soft to stiff). Both constant amplitude mode and AFC mode were used for comparison. A comparison of low- and high-amplitude constant amplitude passes revealed

 areas of positive MV-A dependence (MVs increased with in- creasing A) and negative MV-A dependence (MVs decreased with increasing A). The dependence of roller MVs on A can provide a misleading record of soil stiffness when operating in AFC mode. The positive and negative MV-A dependence was observed during testing and resulted in an artificial and misleading level of variability in recorded soil stiffness. In ad- dition, the roller MV-A dependence can falsely trigger AFC changes in A. This is particularly problematic when roller MVs hover around a target or limit MV. The response dis- tance of AFC tested was found to be approximately 1 m (3 ft), indicating that rollers in AFC mode can respond to relatively local changes in soil conditions. The influence of AFC-based IC on compaction efficiency and uniformity was investigated on granular base material. Two granular base test beds were prepared side by side and compacted with a Bomag Variocontrol system in AFC and constant amplitude modes, respectively. Spot-test measure- ments were obtained at several intermediate compaction passes to assess compaction efficiency and uniformity. Com- paction curves built from spot-test measurement averages revealed similar trends in both test beds with no noticeable difference. The COVs of spot-test measurements were similar after pass 8 for the two test beds, indicating no discernable difference in uniformity. A comparison of COVs from roller MVs recorded during constant amplitude final passes on each test bed showed no difference in uniformity. AFC-based IC did not produce greater compaction or improved uniformity compared to constant amplitude mode compaction for these test bed conditions. 9.6 Relationship Between Roller Measurement Values and Spot-Test Measurements Implementation of roller-integrated CCC into earthwork specifications requires an understanding of relationships between roller MVs and spot-test measurements. A compre- hensive evaluation of five roller-integrated MVs (i.e., MDP, CMV D , E vib , k s , CCV) and 17 different soils was performed. The soils are grouped into three material groups: nongranu- lar subgrade, granular subgrade, and granular subbase/base materials. Roller MVs were obtained from smooth drum and pad foot drum rollers on 60 controlled test beds. The test beds varied in material types, moisture content, and underlying layer support conditions. Roller MVs were obtained for dif- ferent amplitude, frequency, and speed settings. A variety of conventional and mechanistic related in situ test mea- surements [i.e., dry unit weight, CBR, LWD modulus, PLT modulus] and laboratory M r test measurements were used in correlation analyses to MVs. The objectives were to (1) in- vestigate simple linear relationships between roller MVs and various in situ point measurements, (2) identify key factors that influence these relationships, and (3) evaluate multiple regression relationships that consider variations in soil con- ditions and machine operation settings. The results indicated that correlations are possible to dry unit weight, modulus, and CBR with simple linear regres- sion analysis on test beds with homogeneous and relatively stiff underlying layer support conditions and MVs obtained under constant operation settings. A summary of typical ranges of R2 values for modulus, CBR, and dry unit weight measurements for the three referenced material groups is provided in Table 9.1. The wide range in resulting R2 values is attributed to various factors, including sublift heterogeneity, moisture content variation, narrow range of measurements, transverse heterogeneity, and variation in machine operat- ing parameters. High variability in soil properties across the drum width and soil moisture content contributes to scatter in relationships. Averaging measurements across the drum width and incorporating moisture content into multiple re- gression analysis, when statistically significant, can help miti- gate the scatter to some extent. Relatively constant machine operation settings are critical for calibration strips (i.e., con- stant amplitude, frequency, and speed), and correlations are generally better for low-amplitude settings [e.g., 0.7 to 1.1 mm (0.028 to 0.043 in)]. The influence of some of these factors (i.e., soil moisture content, compaction layer lift thickness, underlying layer properties, and machine operation settings) was statistically analyzed using multiple regression analysis. A summary of typical ranges of R2 adj (adjusted for the number of model parameters) values for modulus, CBR, and dry unit weight measurements for the three referenced material groups from multiple regression analysis is provided in Table 9.2. Where heterogeneous conditions were evident below the compac- tion layer, the underlying layer properties (MVs and spot-test measurements) were often found to be statistically significant in the multiple regression model. Regression relationships improved by incorporating the underlying layer properties. Where compaction layer properties were strongly correlated with the underlying layer properties, the compaction layer table 9.1. typical range of R2 values for simple linear regression analysis. Material Modulus CBR γ d Nongranular subgrade 0.1−0.7 0.1−0.7 0.0−0.6 Granular subgrade 0.3−0.7 0.0−0.4 0.1−0.5 Granular subbase/base 0.2−0.8 0.0−0.6 0.0−0.5

  point measurements were not statistically significant in the analysis. Moisture content was found to be significant for two nongranular subgrade layer test beds and one granular base layer test bed. Generally, moisture content was not statisti- cally significant in the regression analysis for most of the test bed studies. Factors contributing to this observation are (1) moisture content did not vary enough over the length of the test strip; (2) spot-test measurements typically only measure moisture content to about 75 mm (3 in) below the surface, while the measurement depth of the roller is much greater; and (3) when correlating with elastic modulus-based spot- test measurements using multiple regression analysis, mois- ture content is co-linear (i.e., highly correlated to in situ mea- surement). Amplitude variation was statistically significant for all cases where minimum amplitude variation of ±0.30 mm was present in the data. An approach to empirically relate laboratory-determined M r for a selected stress condition and roller MVs was pre- sented. The M r values were predicted for in situ w-γ d mea- surements using a w-γ d -M r relationship developed from laboratory testing. Similar to other in situ point measure- ments, the relationships were possible for compaction layer material underlain by homogeneous and relatively stiff sup- port conditions. Heterogeneous supporting layer conditions affected these relationships, and the relationships improved by including parameter values that represent the underlying layer conditions through multiple regression analysis. 9.7 Recommended Specification Options for Earthwork Compaction QA Using Roller- Integrated Continuous Compaction Control Based on a thorough critique of European specifications and practice, a review of previous research, and the findings from the research conducted for this project, recommended specification options for QA of subgrade, subbase, and aggre- gate base course compaction using roller-integrated continu- ous CCC were developed. Six viable QA options are proposed to accommodate the diverse site conditions and agency needs observed across the United States. Many of the recommended options were inspired by current European specifications (summarized in Chapter 2). Additional recommended op- tions stem from the research findings presented in this report. The six recommended QA specification options are summa- rized in Table 9.3 and Figure 9.1. Options are numbered 1, 2a, 2b, 3a, 3b, and 3c and are distinguished by three principal categories. In Option 1, CCC is used to assist in QA, but ac- ceptance is based on spot-test measurements. Acceptance of Options 2a and 2b is based on roller MVs, but initial calibra- tion of roller MV to spot-test measurements is not required. Acceptance of Options 3a, 3b, and 3c is based on achieving a MV-TV determined via various initial calibration techniques. Method-based approaches, such as using GPS positioning and documentation to record the pass sequence, are also viable ap- proaches. Because they do not utilize roller MVs, their imple- mentation is straightforward and not presented here. None of the recommended options constitute performance-based specifications. Further research is required to implement, for example, the findings from Chapter 4 into viable performance- based specifications that can be implemented in practice. The recommended specification options prohibit the use of automatic feedback control IC during QA due to the in- fluence that roller operating parameters have on roller MVs. Automatic feedback control IC may be used during compac- tion operations. The proposed specification is applicable to cohesive and cohesionless soils and aggregate base materials. However, as shown in this study, current roller MVs are less reliable on cohesive soils, and particular attention must be given to soil moisture content. Vibratory and nonvibratory- based roller MVs that can be correlated to soil properties (e.g., density, stiffness, shear strength) are permitted. Each of the recommended specification options can be adopted as the sole method for QA. Alternatively, two or more options can be combined to increase reliability. Uniformity criteria can be added to any of these options. A number of important issues regarding roller-integrated CCC specifications are detailed in Chapter 7 and are worth highlighting here. The implementation of roller-integrated CCC for earthwork QA requires knowledgeable field person- nel. QA personnel must be familiar with the aspects of CCC and IC described in this report and with the CCC and IC equipment on the job site. The specifications require close collaboration between the QA personnel and the contractor (i.e., roller operator). To this end, roller operators must un- derstand the various aspects of measurement passes, roller operating parameters, driving patterns, and so forth, as well as the documentation systems of CCC/IC rollers. table 9.2. typical range of R2adj values for multiple linear regression analysis Material γ d Modulus CBR Nongranular subgrade 0.6–0.8 0.2–0.6 0.3–0.7 Granular subgrade — 0.5–0.7 — Granular subbase/base 0.4–0.8 0.6–0.9 0.4–0.8

 Acceptance testing for all specification options is per- formed on evaluation sections. Proper selection of evalua- tion sections that exhibit consistently distributed heterogene- ity (or homogeneity) in both the longitudinal and transverse directions is critical to successful implementation of CCC specifications. Because roller MVs reflect both the compac- tion lift and sublift properties, it is also critical that the se- lected calibration area (for Option 3) exhibit the same degree of heterogeneity as the evaluation section. Instrumented rollers used for CCC-based QA must meet minimum performance criteria. The documentation sys- tem must display and record roller MVs and their three- dimensional GPS position, vibration amplitude and fre- quency, and travel speed. Data must be easily accessible via the onboard computer and easily exportable for postprocess- ing and record keeping. Onboard computers should perform basic statistical analyses of roller MV data and operational pa- rameters. The instrumented roller must demonstrate a mini- mum level of repeatability and reporting position accuracy. 9.8 Implementation of Specification Options: Case Studies The recommended specification options were imple- mented during field testing in Colorado, Florida, North Car- olina, and Minnesota. The case studies presented included compaction and roller-integrated CCC QA of granular sub- grade, nongranular subgrade, granular subbase, and aggre- gate base material. Multiple specification options were inves- tigated as part of each case study, thus enabling the direct comparison of option strengths and limitations. Compac- tion QA acceptance testing was conducted by the project QA personnel using criteria based on dry density requirements (γ d -TV), moisture requirements (w-TV), and/or static proof rolling. For some case studies, comparisons with existing QA approaches were made. Implementation of the recommended specifications al- lowed a direct comparison of roller-based CCC options with each other and with existing (i.e., random spot-test driven) QA practice. Specification Option 1 requires minimal changes to typical existing QA practices. Rather than selecting random points for spot testing, QA inspectors use the roller MV data map to identify the weakest area(s). Depending on the num- ber of weakest areas identified, the frequency of spot testing may be increased compared to current practice. According to Option 1, if the roller-identified weakest area(s) meet ac- ceptance, the rest of the evaluation section meets acceptance. However, given that testing locations are informed rather than random, requiring the weakest zones to meet 100% of the preexisting QA-TV may be more stringent than current random selection spot testing. Reducing the QA-TV may be more appropriate. Specification Options 2(a, b) and 3(a, b, c) require modifi- cation to current QA practices in that spot-test measurements do not form the basis for QA. Rather, acceptance is granted based on roller MVs. In Option 2a, acceptance is granted when the percentage change in the mean roller MV from pass to pass falls below a preset threshold. In the case studies presented, Option 2a appeared to be less stringent than current practice. It may be desirable to implement Option 2a in conjunction table 9.3. Summary of roller-integrated CCC specification options. Roller-Integrated CCC QA Option Target Measurement Value (MV-TV) Acceptance Criteria Option 1: Spot testing of roller-informed weakest area(s) Not required Spot-test measurements in roller-identified weakest area(s) satisfy contract spot-test measurement requirements (QA-TV) Option 2a: Monitoring percentage change in mean MV Not required Achieving ≤ 5% change in mean MV between consecutive roller passes Option 2b: Monitoring spatial percentage change (%∆) in MVs Not required Achieving the %∆-TV between consecutive passes over a defined percentage of an evaluation section Option 3a: Empirically relating MVs to spot-test measurements Based on correlation of MV to spot-test measurement: MV-TV = MV corresponding to contract QA-TV a Achieving MV-TV over a set percentage of an evaluation section Option 3b: Compaction curve based on MVs MV-TV = mean MV when the increase in pass- to-pass mean MV in a calibration area ≤5% Option 3c: Empirically relating MVs to lab- determined properties (e.g., M r ) Based on correlation of MV to lab soil property: MV-TV = MV corresponding to contract QA-TVb a Assumption is that QA-TV is spot-test-based measurement of density, modulus, etc. b For example, a QA-TV based on M r .

  Figure 9.1. Summary of roller-integrated CCC specification options.

 with Option 1 to improve reliability. Similarly, specification Option 2b uses the percentage change in spatial roller MV data as the basis for QA. One challenge associated with Op- tion 2b is that the method of transforming roller MV data onto a fixed grid to allow spatial comparison is not trivial and reliable, and practical methods do not yet exist. One major challenge to successfully implementing Specifi- cation Option 3(a, b, or c) is ensuring that the calibration area is representative of the evaluation section. Although using a roller MV data map of the evaluation section can aid in se- lecting an appropriate calibration area, this can be logistically challenging on a busy job site. Option 3(a, b, and c) requires a significant initial investment of time and careful, detailed analysis. This analysis is more complex than that currently required for QA, and it is easy to make errors. Accordingly, QA inspectors will need careful training to ensure that they are familiar with both the roller MV systems and the analysis required for the various options. Construction traffic poses a challenge to implementing CCC-based QA. All of the options require careful, repeat- able rolling patterns. Construction traffic, particularly haul trucks moving through the earthwork area, forced less than ideal roller pass patterns. Truck traffic often made it diffi- cult to create uninterrupted and repeatable evaluation sec- tion roller MV maps. Developing the required correlations requires that haul trucks remain outside the calibration area once material has been placed and spot-test measurements are being performed. However, it is common for contractors to utilize haul truck traffic to compact soil. Similarly, it is not uncommon for haul trucks to enter the evaluation section in reverse, deposit their material, and then drive forward out of the area. This forces less than ideal roller pass patterns and creates hazardous conditions for personnel performing spot- test measurements. Performing correlation studies in a designated full-width calibration area requires a change in how the earthwork con- tractor places material. To perform repeatable measurement passes in the evaluation section, the research team had to wait for the earthwork contractor to completely finish haul- ing and placing material in a section. The pace of the pro- duction earthwork placement and compaction frequently limited the time the research team was able to spend in the calibration area. Including the time needed to construct the calibration area, the correlations were developed in approxi- mately 3 to 4 hours, although a time frame of 1 to 2 hours or less would be more consistent with production schedules. In typical production compaction practice, roller compac- tors are used throughout the hauling, placing, and grading operation. Careful planning and cooperation between the contractor and QA personnel, together with modifications in work flow, are critical for successful implementation of CCC- based QA.