Evaluating Applications of Field Spectroscopy Devices to Fingerprint Commonly Used Construction Materials (Phase IV–Implementation) (2014)

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.

CHAPTER 3 Findings and Applications Identification of Participating Agencies The pre-implementation phase of this project started with identifying a group of participating DOTs that would be willing to provide their input on the following three draft AASHTO standard practices that were proposed in the earlier phase of the project (Zofka et al., 2013): 1. Identification of Water Reducing, Accelerating, and Retarding Chemical Admixtures in Fresh Portland Cement Concrete by Attenuated Total Reflection Infrared Spectrometer 2. Determination of Titanium Content in Bridge and Traffic Paints by Field-Portable X-Ray Fluorescence Spectroscopy 3. Evaluation of Oxidation Level of Asphalt Mixtures by Attenuated Total/Diffused Reflection Infrared Spectrometer The research team contacted materials specialists from Connecticut, Maine, New Hampshire, New Jersey, Pennsylvania, and Rhode Island Departments of Transportation (DOTs). Each contact was supplied with the objectives of the project and electronic copies of draft AASHTO standards, and asked to schedule a kick-off meeting. Pennsylvania DOT personnel indicated that they were basically interested in the XRF application, but they did not have the time for full-blown participation in the project. The demonstration sessions for all three methods were held in the headquarters of Connecticut, New Hampshire, Maine, and Rhode Island DOTs. The kick-off meetings revealed different degrees of interest from the potential participants, as shown in Table 3.1. Eventually, Connecticut and Maine stayed on board to provide feedback on the draft AASHTO standards and participate in lab and field testing for all three methods. In general, the method that attracted the most interest from the various agencies was XRF. Portable XRF analyzers may be used for a variety of applications besides paints, including detection of lead (Pb) and arsenic (As) in glass beads, lead in paints and soils, chromium (Cr) in PCC, and other trace metals in a variety of construction materials; thus the acquisition of the equipment can serve multiple purposes simultaneously. In addition, the current QA/QC methods on paint testing do not involve any chemical specifications that can be used to verify that the paint quality is consistent. Thus, interest in this method was expressed by all contacted agencies that responded to the contact attempts. The agency representatives’ reactions to the PCC admixtures specification were, conversely, consistently reluctant. The agency representatives either rejected outright the need for this test or merely consented to try out the method in the field. Specifically, the materials testing professionals felt that the performance-based QA/QC methods for accelerators or retarders were adequate and that early detection of the presence of the compound in the wet 6

admixture would not serve any additional purpose. Because of this limited interest, the team later had difficulty identifying suitable projects on which to conduct the field testing. Finally, the recycled asphalt pavement (RAP) analysis method was received with markedly different reactions depending on the agency. Both agencies that ended up participating were enthusiastic that such a method could become available and expressed an urgent need to have a means of evaluating the RAP content and the degree of oxidation in asphalt. Conversely, other DOTs had no interest in trying the method and found no usefulness in the application. The reactions to the method reflected to some degree the extent of RAP application in the particular state and the views of the asphalt-related personnel on the nature and properties of RAP. On the basis of this feedback, it is considered that the XRF method is likely to be more attractive to state DOTs, contractors, and manufacturers as a quality control tool. The ATR method for PCC does not seem to hold much promise except in states where portland cement concrete is more widely used, which was not the case with the participating DOTs. The RAP method requires additional development and education of professionals to become a viable method. However, the research team concludes that there is substantial need for the method in areas with intensive RAP use. It should be noted that these conclusions hold true for the more common paradigm of design-bid-build, in which the DOTs are responsible to ensure that their design specs were made. However, the team considers that XRF may also be an attractive method for other project management paradigms, in which it used more as an internal standard QA/QC tool, or to assess the current status of materials status, e.g., in bridges. Additionally, changes in design specifications for the service life of bridges, asphalt surfaces, and other materials, may render the method more or less attractive. Table 3.1. Summary of Agency Participation in Pre-implementation Phase Agency Proposed AASHTO Standard ATR of Admixtures for PCC XRF of Paints ATR of RAP and Oxidized Asphalts Connecticut DOT Actively participated in lab and field testing Actively participated in lab and field testing Actively participated in lab and field testing Maine DOT Actively participated in lab and field testing Actively participated in lab and field testing Actively participated in lab and field testing Rhode Island DOT Not interested Participated in demonstration only Participated in demonstration only New Hampshire DOT Not interested Limited interest Not interested New Jersey DOT Not discussed Unresponsive to contact Not discussed Pennsylvania DOT Not discussed Indicated interest but limited availability Not discussed 7

Revision of Proposed AASHTO Standards The next three subsections provide brief descriptions of the tests, the nature of agencies’ comments, and implemented revisions for each draft AASHTO standard separately. A full description of the standards can be found in Appendix A, whereas the background on methods of spectroscopic evaluation of materials can be found elsewhere (Zofka et al. 2013). It was found that the comments obtained by the materials specialists were of a general nature and not geared toward the particular needs of the local DOT. Thus, it is considered that there is no need for agency-oriented specifications; instead a single improved AASHTO draft was produced by incorporating all comments received. ATR FT-IR of Chemical Admixtures for Concrete The identification of the type of a chemical admixture for portland cement concrete (PCC) using a compact ATR FT-IR spectrometer is based on the qualitative comparison of the infrared absorbance spectrum of a pure admixture sample with the spectrum of the PCC chemical mixture. A pure admixture sample is obtained from the storage or feeding tank in a concrete plant. A pure admixture sample is scanned by ATR FT-IR to obtain its absorbance spectrum. The absorbance spectrum presents peaks at specific frequencies that are considered characteristic of the particular chemical. A fresh PCC sample is scanned by ATR to obtain the spectrum of the mixture. If the characteristic absorption bands of the chemical are observed, then the presence of the chemical can be established. The comments on this proposed method provided by the materials specialists were primarily of an editorial nature. One valuable suggestion was to include a sample spectrum for each type of admixture considered in the method. The draft document was edited accordingly (See Appendix A). XRF of Bridge and Traffic Paints The draft AASHTO XRF standard was reviewed by materials engineers of the Maine and Connecticut DOTs. Comments generally fell under one of three categories: the need for additional safety instructions and documentation regarding the hazards and prevention of exposure to ionizing radiation; not enough specific detail in the field and laboratory testing procedures; and more specificity in quality control and calibration sections. The first issue was addressed by including two additional sections of references to radiation safety literature and relevant regulatory standards. Additional detail was included in the procedure to ensure that specific steps are taken by the XRF operator to mitigate the potential for exposure to either the operator or nearby persons, and to ensure compliance with state and federal occupational safety legislation. Substantial detail was added to the procedure to make the steps easier to follow for a field operator. This included additional steps for selecting suitable testing locations, preparation of samples for lab testing, and the optimal technique for data collection, which is utilized by most portable XRF analyzer manufacturers. In addition, a list of potential vendors and manufacturers 8

of XRF analyzers and related materials was included as an appendix. The quality control sections were updated with specific instructions for instrument calibration and periodic standardization. Instructions for how to recognize as well as mitigate commonly encountered interference effects were updated in the procedure. The full text of the revised standard can be found in Appendix A. ATR of Recycled and Oxidized Asphalts No comments from either Connecticut or Maine DOTs were received on this method. This is attributed to the limited familiarity of materials testing engineers with spectroscopy and also with RAP testing, which is not a standard practice, given the lack of relevant standards. In general, while there is considerable interest in development of testing methods and improved understanding of RAP properties, the method requires additional maturing to reach the stage of widespread application as a QA/QC procedure. It should be noted that Connecticut DOT does not own and, therefore, does not control RAP stockpiles, nor does it specify aging properties of RAP materials. Nevertheless, the asphalt materials specialists expressed their interest in the results of this project. Therefore, the research team proceeded with including samples from RAP stockpiles owned by the major CTDOT contractor, Tilcon Connecticut, Inc., in the testing program. Maine DOT engineers, on the other hand, volunteered large amounts of material from 15 different locations and fully participated in the pilot testing for this method. Determination of Signature Spectra of State-Approved Materials Field testing of state-approved materials required the development of appropriate standards, against which the field samples could be compared. Thus, samples of several material types were obtained from both states and tested in the lab with the appropriate procedures described in Appendix B. ATR of Chemical Admixture to PCC Several samples of PCC chemical admixtures were obtained from CTDOT and Maine DOT. CTDOT shipped samples directly to Connecticut Advanced Pavement Laboratory (CAP Lab) at UConn, where the laboratory testing occurred. In Maine, the admixture samples were first transferred from different locations/plants to the Freeport, Maine, materials laboratory. The research team picked up the samples from Freeport and took them to CAP Lab for testing. Table 3.2 provides details on the manufacturer, type, and sample size of admixtures tested. More information on material composition and properties can be found in the corresponding material safety data sheets and product descriptions available online. To perform an ATR test according to the proposed AASHTO standard, about 1 ml of each admixture was probed from the container and placed on top of the ATR diamond plate. Each sample was scanned for about 1 min and the average of 24 absorbance spectra was recorded. Due to the liquid nature of all admixtures, a maximum testing of 2 or 3 subsamples 9

was sufficient to confirm very high repeatability of the method, as shown in Figure 3.1. This way a signature spectrum was determined for each sample listed in Table 3.2, shown in Appendix C. Table 3.2. Summary of Admixture Samples Tested in CAP Lab Participating Agency Admix- ture Name Admixture Type Admixture Manufacturer Source/Plant/ Location Sample Size Connecticut DOT Zyla 630 Water Reducer Type ASTM C494 Type A and D Grace Concrete Products Rocky Hill Lab 64 oz. Zyla R Water Reducer and Retarder ASTM C494 Type B and D Grace Concrete Products Rocky Hill Lab 64 oz. Maine DOT ADVA 140 High Range Water Reducer ASTM C494 Type A and F Grace Concrete Products Sargent Materials, Hermon, Maine 32 oz. Darex II AEA Air Entrainer ASTM C260 Grace Concrete Products Sargent Materials, Hermon, Maine 32 oz. Delvo stabilizer Water Reducer and Retarder ASTM C494 Type B and D BASF Dragon Products Co., Portland, Maine 32 oz. Glenium 7500 High Range Water Reducer ASTM C494 Type A and F BASF Dragon Products Co., Portland, Maine 32 oz. Micro Air Air Entrainer ASTM C260 BASF Dragon Products Co., Portland, Maine 32 oz. Polar Set Accelerator ASTM C494 Type C W.R. Grace & Co. Sargent Materials, Hermon, Maine 32 oz. Pozzolith 100XR Water Reducer and Retarder ASTM C494 Type B and D BASF Dragon Products Co., Portland, Maine 32 oz. Recover Retarder ASTM C494 Type D Grace Concrete Products Sargent Materials, Hermon, Maine 32 oz. Retarder Unknown Unknown Sargent Materials, Hermon, Maine 32 oz. Rheocrete CNI Accelerator/Corrosion inhibitor ASTM C494 Type C BASF Dragon Products Co., Portland, Maine 32 oz. 10

Figure 3.1. ATR absorbance spectra of the CTDOT chemical admixtures. XRF of Traffic Paints Initially, Connecticut and Maine DOT provided samples for laboratory testing. Elemental analysis results were obtained for several white and yellow traffic paints, as well as green bridge paints, to establish the chemical composition of the materials used in the field (Table 3.3). The concentrations would serve as a baseline to determine concentrations in the field. Table 3.3. Titanium Concentrations of Liquid Paint Samples as Determined in the Laboratory by XRF Paint Name/Type Color Average Ti Content (%) Standard Deviation (%) ME DOT Corothane Green 8.82 0.13 ME Spec Waterborne TP White 6.50 0.11 ME Spec Waterborne TP Yellow 1.62 0.08 Epoplex LS-50 CTDOT White 28.78 0.03 Epoplex LS-50 CTDOT Yellow 5.36 0.08 DM3062AM CTDOT Yellow 2.05 0.12 Ennis Latex CTY-21-M-2 Yellow 3.65 0.02 Ennis Latex CTW-21-M-1 White 6.77 0.28 11

The effect of dilution on the XRF signal of latex paints was determined by testing samples corresponding to different paint-to-water ratios (Figure 3.2). Each paint type shows a decay-type curve between approximately 1% to 5% titanium (Ti). A linear relationship is seen between the diluted Ti content and the measure Ti content below 1% Ti. Thus, water-diluted paints may be detected through the development of such calibration curves. Figure 3.2. Effect of dilution on the Ti XRF signal of three latex-based traffic paints. ATR of Recycled Asphalt Materials Sample Preparation and Testing In Connecticut, RAP samples were collected from three asphalt plants owned by Tilcon Connecticut, Inc., located in Manchester, New Britain, and Groton. Those locations were chosen to represent different climatic subzones, specifically inland hills and shore plains. It was hypothesized that difference in precipitation and location may affect progress of oxidation and moisture content in RAP during the storage. Other factors considered in this study were RAP source (state versus private projects), RAP gradation (milled versus processed to passing #4 sieve), duration of storage (e.g., 2 years versus 6 years), and location within a stockpile (top/ crust versus inside). Table 3.4 summarizes information on Connecticut DOT RAP samples. In Maine, samples from 12 locations spread over a 250-mi stretch of Interstate 95 between Portsmouth and Houlton were gathered in the Freeport Materials Laboratory and transferred to CAP Lab for testing. The samples differed in gradation (from less than 8% to more than 10% aggregate passing #200 sieve) and variation in asphalt content (from less than 0.3 to greater than 0.5%). Table 3.5 provides a summary of the Maine DOT RAP samples. All RAP samples were stored in sealed plastic bags or tin containers to preserve the original moisture content. Prior to ATR testing, each sample was sieved through #8, #30, and #50 sieves, and 5 subsamples of the fraction between the #30 and #50 sieves (diameter between 12

0.59 mm and 0.297 mm) was tested from each sieved sample. This allowed for maximizing the uniformity of particle size and minimizing the variability of sample composition. Table 3.4. Summary of Connecticut DOT RAP Sample Information RAP Stockpile Location RAP Source Duration of Storage Milled or Processed Top or Inside Groton State 1–2 years Processed Inside Groton Private N/A Processed Inside Groton State 2 years Milled Top Manchester Private N/A Processed Top Manchester Private N/A Processed Inside Manchester State N/A Processed Top Manchester State N/A Processed Inside New Britain State 1–2 years Milled Inside New Britain State 1–2 years Milled Top New Britain State 5–6 years Milled Inside New Britain State 0–1 year Processed Top New Britain State 0–1 year Processed Inside 13

Table 3.5 Summary of Maine DOT RAP Sample Information RAP Stockpile ID Nearby Location Latitude, °N Longitude, °W RAP Classification Passing #200 Standard Deviation in Asphalt Content LAN-HE13-CR- Q Hampden, Maine 44.785 –68.8319 Pure aggregate N/A N/A LAN-SM13-II Smyrna, Maine 46.1301 –68.1579 Class II ≤10% ≤0.5 PII-WE13-I Wells, Maine Wells, Maine Wells, Maine Class I ≤8% ≤0.3 PII-HE13-I Hampden, Maine 44.7783 –68.8395 Class I ≤8% ≤0.3 PII-FF13-II Fairfield, Maine 44.5908 –69.6134 Class II ≤10% ≤0.5 LAN-WB13-II Wetsbrook, Maine 43.7321 –70.3527 Class II ≤10% ≤0.5 PII-PNH13-I Portsmouth, N.H. 43.0466 –70.7794 Class I ≤8% ≤0.3 CMP-SC13-III Dayton, Maine 43.5294 –70.5931 Class III >10% N/A PII-WB13-I Wetsbrook, Maine 43.6742 –70.3320 Class I ≤8% ≤0.3 LAN-PI13-II Presque Isle, Maine 46.7359 –67.9614 Class II ≤10% ≤0.5 PII-PO3-I Poland, Maine 44.0165 –70.3491 Class I ≤8% ≤0.3 LAN-WA13-III Washington, Maine 44.2226 –69.3853 Class III >10% N/A To obtain an infrared (IR) absorbance spectrum, each sample was scanned 24 times on the ATR diamond sampling plate and the average spectrum was recorded. The spectrum was then processed in accordance with the proposed draft AASHTO method and the analysis of oxidation level in terms of carbonyl content and comparison of moisture content was performed. Note that the proposed AASHTO method of measuring oxidation in asphalt mix samples is qualitative by nature and therefore only comparative analysis is provided in this report. Development of absolute standards would require a much larger effort of collecting RAP nationwide, evaluating its properties and developing a database that could provide a basis for comparison. Analysis of Oxidation and Moisture in RAP Samples In Connecticut, the results indicated that it was possible to compare both oxidation and moisture in samples from stockpiles with different storage duration. For example, Figure 3.3 compares the carbonyl and aromatic content in different RAP samples, as determined by the absorbance peaks around 1700 cm–1 and 1600 cm–1, respectively. An increase is evident in signals from the 14

carbonyl and aromatic groups with storage time. The signal from the water presence at around 3350 cm–1 indicates higher moisture content for the 5-year-old stockpile. It is assumed that the seasonal variation of moisture occurs simultaneously in both piles because of their proximity (within one plant). Figure 3.3. Comparison of moisture (left) and oxidation and hardening (right) in RAP stockpiles after 1 to 6 years of storage. The main challenge in testing RAP or any asphalt mix sample is a relatively high (up to 25% for five subsamples tested in this study) standard deviation from the mean. Ultimately, after gaining experience with the testing method, the analyst may conduct 10 analyses and choose the five most representative of the bulk material to reduce variability to as low as 10% to 15%. Considering the very short duration of a test (less than 1 min), reliable results for a sample prepared as a single batch can be obtained within 10 to 15 min, which complies with the objective of this research. The limiting factor is that an experienced technician is required to judge the quality and representativeness of the spectra. In Maine, the following two factors of RAP oxidation were investigated: (1) effect of location and (2) effect of RAP class/gradation/asphalt content. A limitation to studying the influence of other factors was that the age of RAP and duration of its storage in a stockpile were unknown. Interestingly, it was found that on average, oxidation in RAP samples was decreasing with latitude, as shown in Figure 3.4. One possible explanation of this trend is the use of binder heavily modified with styrene–butadiene–styrene (SBS) in northern latitudes, such as PG 76-28, which is significantly stiffer than, for example, PG 58-28 and PG 64-22. The stiffer binders are known to oxidize at a slower rate. As far as RAP classification is concerned, the average oxidation was higher in Class I, followed by Classes III and II, in that order (Figure 3.5). A statistical significance of this phenomenon should be further investigated on a larger set of samples. In summary, however, the ATR method revealed considerable variability in oxidation levels of RAP from different projects, which further indicates the need of controlling this parameter, especially when producing mixes with high RAP content. 15

Figure 3.4. Effect of location on RAP oxidation in Maine. Figure 3.5. Summary of oxidation in different RAP classes. Pilot Field-Test Results ATR of Admixtures No field tests were performed for this method because no suitable project could be identified. This is related to the limited need and interest of the DOTs for the method, given that these admixtures could only be tested for QA/QC in the designated tank at the manufacturer or the concrete ready-mix plant. In general, a stable chemical composition of admixtures and their liquid nature renders them suitable for QA/QC analysis; however, the point of application is the manufacturing location and/or material source, rather than the construction site. 16

XRF of Bridge and Traffic Paints in Maine In situ XRF testing was performed in collaboration with the Maine Department of Transportation. The first field test was conducted on a bridge overpassing I-95 one mi north of Exit 174. The site is located on Hinkley Hill Road in Herman, Maine, approximately 8 mi west- southwest of central Bangor. X-ray data were collected by directly applying the XRF analyzer on the green-painted girders on the underside of the bridge. Points of measurement were accessed by means of a boom lift stationed on the right southbound lane of I-95. Pictures of the field- testing processes are shown in Appendix C. The most abundant detectable elements were Ti, iron (Fe,) and zinc (Zn) (Table 3.6). Smaller concentrations of zirconium (Zr) and Pb were detected at fractional mass-percentages. Comparison to the lab-based results revealed little variation in the Ti content between the sampling techniques. The block samples and liquid samples had slightly lower and higher Ti concentrations, respectively, compared to the bride girder. This similarity suggests that the Ti content of painted girders measured in situ is generally representative of the lab-based characterization. In contrast, substantial variation is seen among the elemental concentrations of Fe, Zn, and Pb. This is suggestive of interference from elements present in the substrate materials which are absent from the pure paint. Table 3.6. Elemental Composition of Bridge Paint as Detected by XRF for Liquid Samples, Coated Metal Wedge Samples, and the Painted Bridge Girder Element Laboratory Ti Content (%) Field Ti (%) Liquid Paint std Painted Wedges std Bridge std Ti 8.82 0.13 2.67 2.67 9.21 0.29 Fe 0.02 0.01 13.30 13.30 17.83 3.68 Zn 0.01 0.01 ND – 34.20 4.85 Pb ND – ND – 0.02 0.01 The additional elements detected in the field may be readily explained by the chemical composition of the underlying steel and residual coatings. Lead-based paints were used extensively for corrosion inhibition on steel bridges prior to regulatory changes in the 1970s. Thus, the Pb detected by the XRF likely resides in residual paint layers from the previously applied coatings. Measurements performed on an unpainted section of a girder showed an average of 0.1% Pb, indicating that approximately 20% of the Pb remained after the coating. At about 17% and 37%, Fe and Zn were the most abundant elements detected, respectively. Lab- based tests on liquid paint samples showed that the paint did not contain appreciable amounts of Fe (<0.04%) or Zn (<0.01%). This indicates that the X-rays penetrated depths beyond the paint coating into the girder substrate material, which is most certainly galvanized steel. The galvanization process produces a thick protective layer of Zn on steel, an alloy of primarily Fe and carbon. Thus, the detection of Fe further indicates that the XRF probed the bulk material at depths beyond the exterior paint and Zn coatings. 17

Collectively, these results show that XRF is effective for field testing the Ti content of bridge paint, in that in situ measurements closely matched lab measurements. The wide variation exhibited by other elements is suggestive of potential for interference effects depending on the substrate. It is critical, therefore, to test an uncoated sample to verify that the measured concentrations are quantitatively representative of the true paint concentration, since it would be difficult to separate the fluorescence contributions from multiple sources. If the paint is titanium- based, which is the case for most brands, then, Ti is a suitable element for QA/QC testing of the paint itself, given that there are no substantial Ti contents in most materials where paint is applied. Field measurements were also performed on strips of traffic paint dried on Jersey barriers situated behind the Bangor DOT office. The results are quite close to the characteristic concentrations obtained for liquid samples in the laboratory (Figure 3.7). The concentrations were slightly higher in the field, which is likely due to drying. Table 3.7. Titanium Content in Painted Lines at DOT, Bangor, Maine Ti Content (%) White Yellow Sample 1 7.36 1.86 Sample 2 7.47 1.93 Sample 3 7.89 1.82 Average 7.57 1.87 Std 0.28 0.06 XRF Testing of Traffic Paints in Connecticut Field XRF measurements were also performed on painted white and yellow lines in conjunction with the ConnecticutDOT at four locations in southeastern Connecticut: the Exit 9 off-ramp on Route 9 in Higganum, Conn., Route 154 in Old Saybrook, Conn., and Exits 64 and 66 on Interstate 95 in Westbrook, Conn. (See Figure 3.8.) The only major difference in locations was that the Route 154 site had been freshly painted one or two days prior to measurement. The Ti contents of the white paints tested at both I-95 locations are nearly identical, which indicates a similar paint mix, paint type, or application technique. The Ti content in the yellow paint was slightly lower at Exit 64. The Route 9 site had the highest Ti content at 22%, which is comparable to the high-Ti Epoplex paint characterized in the lab. This site also produced a higher standard deviation, which is either a consequence of the higher Ti content or an inconsistency in the paint thickness. Freshly applied paint tested at the Old Saybrook site had the lowest Ti concentration (9%). Unfortunately, the XRF device could not be used by any DOT workers in either Maine or Connecticut because of the OSHA guidelines regarding the use of X-ray devices. Nevertheless, the measurement process and draft standard were demonstrated and workers were present for all field tests. The response and feedback were positive in both states. Most comments were centered around the simplicity and ease of use of the device, as well as the significant potential 18

for quality control applications. In Connecticut, interest was expressed in linking XRF data to thickness measurements and other techniques for the purposes of enhancing quality control and optimizing the paint application procedure. Table 3.8. Titanium Concentrations in Pavement Markings as Determined by XRF Location in Connecticut Ti Content (%) White Yellow Exit 66, I-95, Westbrook 12.71 3.44 12.89 4.25 11.25 3.99 12.14 3.51 Average 12.25 3.80 Std 0.74 0.39 Exit 64, I-95, Westbrook 11.75 2.08 12.17 2.90 11.43 2.32 Average 11.78 2.43 Std 0.37 0.42 Route 154, Old Saybrook 7.80 10.79 9.04 Average 9.21 Std 1.50 Exit 9, Route 9, Higganum 22.19 2.83 26.11 2.75 19.93 2.26 18.57 3.22 Average 21.70 2.76 Std 3.30 0.40 ATR of RAP and Oxidized Asphalt Pilot field testing of aged pavement surfaces was performed in Maine. Three pavement locations with a similar pavement structure and mix properties yet of different age (0.5, 5, and 11 years) were selected by Maine DOT from the thin overlay preservation projects. The powdered samples of asphalt were collected by drilling and tested on-site using the portable ATR spectrometer installed on the back of a van. One sample from each pavement section was collected and three probes from each sample bag were tested. Note that the whole operation occurred at an ambient temperature of 35°F, which is much lower than the minimum of 64°F prescribed by the manufacturer. Nevertheless, the instrument did not exhibit any operation problems, nor was the data processing by a laptop computer affected by the low temperature. The results of testing are presented in terms of the carbonyl index, which is determined as the ratio of the absorbance peak at 1700 cm-1 over the peak value at 2920 cm-1. Figure 3.6 plots the resulting carbonyl index as a function of pavement age, showing a fairly significant (R2 = 19

0.46) linear trend of increase in oxidation with pavement age. This result is aligned with results on pavement aging reported for Connecticut and Rhode Island pavements in the previous phases of the project. For example, in Connecticut, pavement sections with binder grade of PG 64-XX exhibited intercept of 0.003 and slope of 0.003. Figure 3.6. Trend of the carbonyl index determined by ATR analysis with the pavement age in Maine DOT field tests. During the field testing of pavement surfaces, a material engineer from Maine DOT was asked to collect a sample and perform an ATR test. It was confirmed that no special training is required for sample collection, whereas a basic training on the instrument software interface was needed to familiarize an operator with basic spectrum manipulation options. 20