Nondestructive Testing to Identify Concrete Bridge Deck Deterioration (2012)

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.

68 C h a p t e r 7 The performance of NDT technologies was generally evalu- ated from the perspective of five performance measures: accuracy, precision (repeatability), ease of use, speed, and cost. The field validation testing focused mainly on the evalu- ation of speed, precision, ease of use, and cost, because the evaluation of accuracy was limited by the extent of ground truth information. Evaluation of the technologies’ accuracy in detecting deterioration, however, was the primary objec- tive of the laboratory testing. assessment of NDt technologies The criteria for the evaluation and ranking of the NDT methods and devices were discussed in Chapter 4. Those criteria were applied, with some minor adjustments, using the quantitative results obtained from the field and laboratory studies. Accuracy The accuracy was judged on the basis of the following three criteria defined in Table 4.3: (1) detectability extent, (2) detect- ability threshold, and (3) severity of deterioration. Detect- ability is the most fundamental parameter for the evaluation of any NDT technology. If a certain defect cannot be detected by a given technology, the other four performance measures are meaningless. In principle, the detectability can be trans- lated into the ability of a given technology to locate a given defect as accurately as possible. At the same time, the same technology should not report an intact location as defective. In other words, the number of false-positive and false-negative results should be minimal. Delamination Five groups of technologies were evaluated for the detection of delamination (see Table 6.4). For delamination detection, the detectability extent was judged by how accurately each technology detected the area and existence of the delami- nated areas. The detectability threshold was assessed by the smallest delaminated area that a technology could detect. The severity of the deterioration was judged by how well a tech- nology could be delineated among advanced stages of delam- ination versus the onset of delamination and whether it can distinguish between a deep or shallow delamination. Based on the criteria defined in Table 4.3 and the preceding expla- nation, the accuracy of the five technologies is presented in Table 7.1. When multiple vendors used the same technology, the average and maximum points earned are reported. The average point can be considered as the status of the state of the practice at this time and the maximum point corresponds to the state of the art. In general, the IE method is the most accurate technology with an average point of 2.8/5. The infrared thermography technology and the GPR were reasonably accurate as well. The main concern with the infrared thermography technol- ogy was that it was very sensitive to the environmental condi- tions and that there is a small window during the day during which the technology works well. The ranking of the GPR would have improved if the delaminated areas were moisture filled and chloride filled. The materials used in the laboratory experiments to simulate the delamination were more repre- sentative of the air-filled delamination. As reflected in Chap- ter 6, chain dragging was not very successful for detecting the delaminated areas that were small or deep. None of the ven- dors conducted the impulse response tests on the laboratory specimens; as such, the subsequent ranking of the method was primarily based on the work done on the Virginia bridge. Rebar Corrosion Three technologies were evaluated for rebar corrosion. The evaluation criterion was based on the ability of the participants to distinguish the difference between the area constructed with Evaluation and Ranking of NDT for Condition Assessment of Bridge Decks

69 Table 7.1. Grading of NDT Technologies Based on Accuracy Defect Technology Participant Device Performance Parameters for Accuracy Grade Average Grade Maximum Grade Detectability Extent (WF = 0.3) Detectability Threshold (WF = 0.3) Severity of Deterioration (WF = 0.4) Delamination GPR 8 Air coupled 3 3 1 2.2 1.7 2.2 4 Ground coupled 3 3 1 2.2 5 Ground coupled 1 3 1 1.0 9 Ground coupled 3 3 1 2.2 IE 4 Scanning system 3 3 3 3.0 2.8 3.4 6 Air coupled 5 5 1 3.4 7 Air coupled 3 3 1 2.2 1 Scanning system 3 3 1 2.2 9 Scanning system 3 3 3 3.0 IE-USW 9 Stationary 3 5 1 2.8 2.8 2.8 Infrared 2 Handheld camera 3 3 1 2.2 2.2 2.2 Chain dragging 9 NA 3 1 1 1.6 1.6 1.6 Rebar corrosion GPM-HCP 4 Stationary 3 1 3 2.4 2.4 2.4 GPR 1 Ground coupled 1 3 1 1.6 1.6 1.6 5 Ground coupled 1 3 1 1.6 HCP 9 Stationary 3 3 1 2.2 2.2 2.2 Crack depth SASW 1 Stationary 3 1 1 1.6 2.3 3.0 4 Stationary 3 3 3 3.0 SWT 7 Stationary 3 3 3 3.0 3.0 3.0 TOFD 7 Stationary 3 1 1 1.6 1.6 1.6 Concrete degradation USW 9 Stationary 3 3 5 3.8 3.8 3.8

70 corroded rebar and the rest of the deck that was built with new rebar. The GPM, HCP, and ER technologies were marginally successful in reporting changes in parts of the deck that were constructed with the corroded rebar. However, it should be also mentioned that the three technologies were not developed to detect or measure the degree of rebar corrosion. Instead, they are used to measure corrosion activity, corrosion rate, or to describe the corrosive environment of concrete. One of the limitations of the reviewed and evaluated tech- nologies is the assessment of the activity and rate of corrosion when epoxy-coated rebars are used in the deck. Although the measurements have been conducted and reported using GPM and HCP, no clear guidelines regarding the interpreta- tion of the measurement results can be provided. The charac- terization of a corrosive environment using ER, however, is not affected by the rebar type in the deck. Depth of Vertical Crack The evaluation of the three sonic methods that were used to detect the depth of cracks was rather straightforward because the depth of the embedded cracks was known, as discussed in Chapter 6. The most promising methods were SWT and SASW used by one of the participants. Because the cracks were evident from the surface, a detectability extent of 3 was given to all methods. Concrete Degradation Only one participant reported information about concrete degradation. As reflected in Table 7.1 and based on tests of the Virginia bridge, the USW method was successful in quantify- ing the degradation of concrete. Strictly speaking, the limita- tion of that method is that the delaminated and cracked concrete show as degraded concrete as well. Repeatability One of the challenges in making an objective comparison between technologies and participants is that the repeatabil- ity of the presented results needed to be made for different data or results “levels.” A level indicates whether the repeated data are just collected, are collected and analyzed, or are analyzed and interpreted. For example, HCP and ER do not require any data reduction, and the condition maps are gen- erated from the raw collected data. However, GPR and IE results must be analyzed before generating the contour maps. One approach to measure the technology repeatability was by using the coefficient of variation (CV). For each individual test, CV at every test point was obtained by calculating the standard deviation s of the three runs, divided by their cor- responding mean value µ (CV = s/µ). The upper and lower bounds of CVs with a 95% confidence interval (CI) were cal- culated for each test by adding and subtracting the CI to and from the mean CV. Figure 7.1 depicts the CV, and the upper and lower bounds for each test. Figure 7.1a shows the repeatability results for the technologies detecting delamination, and Figure 7.1b demonstrates the results for technologies detecting and describing corrosion activity or environment. As can be seen from Figure 7.1a, the IE system of Participant 1 has the lowest CV (better repeatability), followed by Participants 7, 3, and 4. It should be mentioned that the data used in the repeatability study varied between the participants who presented the simple analyzed data and those who presented the interpreted data. Given that the CVs are all rather small (<0.25), one can argue that all of them have an acceptable repeatability. The same argument is valid for the technologies detecting corro- sion; electrical resistivity has the lowest CV followed by GPR and HCP (Figure 7.1b). Electrical resistivity and GPR have a smaller half-width compared with HCP, which shows that these two tests are more repeatable than HCP. However, the final repeatability grading of technologies could not be based on CV because some of the participants submitted their raw data for which it was not possible to calculate CV. This is the reason some of the participant results are missing in Fig- ure 7.1. Therefore, grading based on the provided graphical presentations of the data and results, although somewhat subjective, was selected as a better approach to repeatability evaluation. Unlike the other performance measures, repeatability is hard to quantify for infrared thermography. However, paying attention to the infrared thermography concept helps to clarify the issue. To detect damage in the deck, infrared thermography uses relative values (i.e., the temperature difference between the deteriorated sections and those that can be described to be in sound condition). The stronger the contrast is, the more pronounced the deterioration (delamination) will be. Thus, as long as the heat conduction conditions are the same, it will be repeatable. The effect of debris on the deck, vehicles and people casting shadows, markings on the deck, time of the day, and so forth, however, also need to be taken into account. The grades for repeatability are provided in the summary table (Table 7.7). Speed Speed as a performance measure has two main components: speed of data collection and speed of data analysis and inter- pretation. For most agencies, speed of data collection is more important because of the cost of traffic control and losses and inconveniences associated with traffic interruptions. The data collection speed of the participating technologies was evaluated on the basis of the records taken during the con-

71 duct of the field validation testing. The data analysis and interpretation speed were graded on the basis of the informa- tion provided by the participants. Speed of Data Collection and Coverage Some technologies collect data in a continuous manner, whereas other technologies are spot measurements. There- fore, the speed of data collection is expressed in terms of the production rate, which is the area coverage per hour needed for high–density data coverage. The high density of data coverage in this case is the data collected on a 2-ft × 2-ft grid. Therefore, the speed was calculated by using the area cover- age per hour for both continuous and point measurements. The time taken to perform the repeatability measurements was excluded in the speed calculation. A grade of 1 to 5 was assigned to each technology, with 1 being the least favorable and 5 being the most favorable (the fastest technology in terms of speed of data collection). The grading was assigned as follows: • 5: Area coverage greater than 2,000 ft2/h; • 4: Area coverage between 1,500 and 2,000 ft2/h; • 3: Area coverage between 1,000 and 1,500 ft2/h; • 2: Area coverage between 500 and 1,000 ft2/h; and • 1: Area coverage less than 500 ft2/h. Speed of Data Analysis and Interpretation Data analysis is defined as the processing of raw data collected by the device and includes preprocessing (e.g., background removal); data analysis and presentation, whether this is done Figure 7.1. Repeatability for detection of delamination (a) and characterization of corrosion activity or corrosive environment (b). (a) (b) 0 0.05 0.1 0.15 0.2 0.25 Co ef fic ie nt o f V ar ia nc e Technologies Corrosion Upper Bound Mean CV Lower Bound ER (8) HCP (8) GPR (8) 0 0.05 0.1 0.15 0.2 0.25 0.3 Co ef fic ie nt o f V ar ia nc e Technologies Delamination Upper Bound Mean CV Lower Bound IE (7) IR (3) IE (1) IE (9)

72 by a single parameter or a graphical output; and data inter- pretation. The grades are defined on the basis of the produc- tion rate. Again, the equivalency is expressed in terms of the number of spot measurements analyzed and interpreted. The speed evaluation results for each participant are presented in Table 7.2. The grades were assigned as follows: • 5: Data analysis for more than 1,000 ft2/h; • 3: Data analysis between 500 and 1,000 ft2/h; • 1: Data analysis less than 500 ft2/h. The overall grade was obtained by assigning a weight factor of 0.6 to the speed of data collection and a weight factor of 0.4 to the speed of data analysis and interpretation. The overall grades for each particular technology are calcu- lated for the ultimate ranking of the technology. The results for both the data collection and data analysis are summarized in Table 7.3. The average grade is the mean of all grades cal- culated for all the participants. A grade of 1, 3, or 5 was assigned to each technology, with 1 being the least favorable and 5 being the most favorable (the slowest technology in terms of speed of data analysis and interpretation). Infrared thermography, ground-penetrating radar, electri- cal resistivity, half-cell potential, and impulse response can be conducted at a higher speed rate than the other technologies. Chain dragging and hammer sounding has a medium speed, and galvanostatic pulse measurement, ultrasonic surface waves, and impact echo are of the lowest speed. Table 7.2. Grading of the NDT Technologies Based on Speed Speed (ft2/h) (Raw Data) Speed (Grade) Overall GradeTechnology Participant Data Collection Data Analysis Data Collection (WF = 0.6) Data Analysis (WF = 0.4) Infrared 2 1,750 1,008 4 5 4.4 GPR (Aladdin) 4/5 2,931 Information not provided 5 3 4.2 GPR 8 2,871 600 5 3 4.2 GPR 9 2,313 800 5 3 4.2 GPR 4 2,032 Information not provided 5 3 4.2 Electrical resistivity 9 1,415 1,008 3 5 3.8 Infrared 10 1,227 Information not provided 3 5 3.8 Half-cell potential 9 1,156 1,008 3 5 3.8 Impulse response 3 1,217 1,008 3 5 3.8 GPR (hand cart) 1 2,188 252 5 1 3.4 GPR 5 2,060 120 5 1 3.4 Chain dragging and hammer sounding 9 802 1,008 2 5 3.2 IE-SASW (surface wave) 4 1,654 Information not provided 4 1 2.8 IE 4 1,569 Information not provided 4 1 2.8 IE 1 1,551 168 4 1 2.8 Galvanostatic pulse 4 517 Information not provided 2 3 2.4 PSPA 9 516 630 2 3 2.4 IE 9 1,018 336 3 1 2.2 IE 6 797 336 2 1 1.6 Air-coupled IE 7 575 336 2 1 1.6 USW (for vertical cracks) 4 na na na na na Vertical cracks (surface wave: USW) 7 na na na na na Note: WF = weight factor; PSPA = portable seismic property analyzer; na = not applicable.

73 Ease of Use The following parameters were considered in evaluating the ease of use of each technology: expertise in data collection, number of operators, ease of maneuvering, physical effort for the setup, expertise in data analysis, and potential for automation. The grades for some of the parameters are based on observations at sites where the tests were being performed, including expertise in data collection, number of operators, ease of maneuvering, and physical effort for setup and movement. Grading for “expertise in data analy- sis” and “potential for automation” is based on both the information provided by the participants and the judgment of the research team, which is based on the submitted reports from participants (Table 7.4). The results are summarized in Table 7.5. Table 7.3. Speed Technology Number of Participants Maximum Grade Average Grade Infrared 2 4.4 4.1 GPR 5 4.2 3.9 ER 1 3.8 3.8 HCP 1 3.8 3.8 Impulse response 1 3.8 3.8 Chain dragging 1 3.2 3.2 GPM 1 2.4 2.4 USW 1 2.4 2.4 IE 6 2.8 2.3 Table 7.4. Grading of NDT Technologies Based on Ease of Use Technology Participant Data Collection Data Analysis Potential for Automation Overall Index for Ease of UseWF = 0.45 WF = 0.4 WF = 0.15 Infrared 10 3.7 5 3 4.1 Chain dragging/hammer sounding 9 3.4 5 3 4.0 Infrared 2 3.2 5 3 3.9 Resistivity 9 3.7 3 5 3.6 HCP 9 3.2 3 5 3.4 Impulse response 3 2.3 3 3 2.7 GPM 4 2.1 3 3 2.6 IE 9 3.0 1 5 2.5 GPR 9 3.2 1 3 2.3 GPR 1 3.2 1 3 2.3 GPR 5 3.2 1 3 2.3 GPR 4 3.2 1 3 2.3 IE-SASW (surface waves) 4 3.2 1 3 2.3 Acoustic sounding 4 3.2 1 3 2.3 USW 9 3.2 1 3 2.3 GPR 4/5 2.8 1 3 2.1 IE 1 2.8 1 3 2.1 GPR 8 2.3 1 3 1.9 Air-coupled IE 7 2.3 1 3 1.9 USW (for vertical cracks) 4 1.9 1 1 1.4 Vertical cracks (USW) 7 1.9 1 1 1.4 IE 6 1.0 1 3 1.3 Note: WF = weight factor.

74 Cost The cost of an NDT method includes the following: • Cost of data collection. The cost of data collection is defined as the overall cost based on the workforce size, expertise level, and time needed to collect data. • Cost of data analysis and interpretation. The cost of data analysis and interpretation is defined as the overall cost, based on the workforce expertise level and time needed to analyze and interpret. • Cost of equipment, supplies, and equipment maintenance. • Cost of traffic control. The cost needed to provide a safe work area on the bridge during data collection. To evaluate the cost-effectiveness of each technology, each participant was asked to provide a cost estimate for two hypo- thetical bridges that have a deck area of 5,000 ft2 and 10,000 ft2. The two bridges would be located within 100 miles of the par- ticipant’s office. Because this was assumed to be a production- type job, the cost of the equipment, supplies, maintenance, and so forth were included in the quoted cost. Consequently, the cost of the equipment is not considered as an independent parameter in the cost evaluation. The pricing provided by the participants includes the cost associated with data collection, data reduction, and data interpretation. Figure 7.2 represents the cost per square foot as reported by the participants. As can be seen from Figure 7.2, a few of the participants provided quotes that were unrealistically low or high. The reason is that some of the participants are either researchers, equipment vendors, or manufacturers who rarely work on production-level jobs. The other, and well-expected, observa- tion is that as the bridge size increased, the cost of the survey per unit area decreased. The participants and technologies were divided into five categories: (1) radar (ground penetrating and air coupled); (2) acoustic (impact echo, impulse response, ultrasonic sur- face waves); (3) infrared thermography; (4) chain dragging and hammer sounding; and (5) electrochemical techniques (half-cell potential, electrical resistivity). The average survey cost per square foot was then calculated for each category. Figure 7.3 depicts the cost per square foot of the deck for each category. The data presented in Figure 7.3 do not include costs asso- ciated with traffic control, which can often be substantial. The speed of surveying has a direct effect on the amount of time Table 7.5. Ease of Use Technology Number of Participants Maximum Grade Average Grade Infrared 2 4.1 4.0 Chain dragging 1 4.0 4.0 ER 1 3.6 3.6 HCP 1 3.4 3.4 Impulse response 1 2.7 2.7 GPM 1 2.6 2.6 GPR 5 2.3 2.2 IE 6 2.5 2.1 USW 1 1.4 1.4 Figure 7.2. Cost per square foot for the assumed surveys on 5,000-ft2 and 10,000-ft2 bridge decks (excluding traffic control). $- 5 8 1 7 9-I E 9-U SW 9-C ha in Dr ag 2 9-H CP 9-E R 4-I E 6 3 9-G PR $0.20 $0.40 $0.60 $0.80 $1.00 $1.20 $1.40 $1.60 $1.80 $2.00 Su rv ey C os t/ 2 Parcipant 5,000 2 bridge 10,000 2 bridge Radar Acousc Electrochemical In fr ar ed th er m og ra ph y

75 traffic control is needed. On the basis of the speed calculated for each technique during the field validation testing in Hay- market, Virginia (Table 7.2), the total time that each partici- pant would need to survey the two hypothetical bridges was calculated. A full day of traffic control was considered to be 6 h: from 9 a.m. to 3 p.m. Because the typical minimum time frame for traffic control is half a day, the survey times were rounded to half-day increments: less than 3 h (0.5 day), between 3 and 6 h (1 day), between 6 and 9 h (1.5 days), and so forth. Figure 7.4 depicts the number of days needed for each partici- pant to survey the bridge. A typical full day of traffic control is assumed to cost $2,000. For the air-coupled GPR, the assumed cost is $1,000, because an air-coupled GPR does not require an entire lane to be closed to perform the testing. Instead, an attenuator truck following the vehicle on which the air-coupled GPR is mounted would suffice in many situations. Figure 7.5 com- pares the overall cost of testing per square foot in the two situations: with and without traffic control. Ground-penetrating radar is the fastest technology on the deck, and traffic control only adds about 27% to the overall cost. Conversely, in the case of the least expensive technologies such as chain dragging and hammer sounding and electro- chemical methods, the traffic control significantly adds to the final cost: 384% and 205%, respectively. However, chain drag- ging and hammer sounding are the only techniques whose Figure 7.3. Cost per square foot for each technology category (excluding traffic control). $- $0.10 $0.20 $0.30 $0.40 $0.50 $0.60 $0.70 $0.80 Radar Acous c Chain Dragging and Hammer Sounding Infrared Electrochemical O ve ra ll Co st / 2 (e xc lu di ng th e tr affi c co nt ro l) 5,000 €2 Bridge 10,000 €2 Bridge Figure 7.4. Number of days to survey the bridge. $- $0.50 $1.00 $1.50 $2.00 $2.50 O ve ra ll Co st / 2 5,000 2 Bridge 10,000 2 Bridge Radar Acousc Electrochemical In fr ar ed t he rm og ra ph y 5 8 1 7 9-I E 9-U SW 9-C ha in Dr ag 2 9-H CP 9-E R 4-I E 6 3 9-G PR Parcipant

76 speed changes with the level of deterioration because there is no postprocessing involved and testing and data analysis are performed simultaneously on the bridge. The deck of the Haymarket Bridge in Virginia, which was used in the field validation phase of the project, has significant signs of dete- rioration. Therefore, the testing took a longer time than usual, which brought the cost of the chain dragging and hammer sounding to the same level as the infrared thermography. The grading of the technologies is based on the unit cost for the hypothetical evaluation of a bridge deck of 5,000 ft2 (Table 7.6). The technologies costing less than $0.5/ft2 were assigned Grade 5, and the grade was decreased one point for every additional $0.25/ft2. Summary Grades The grades for all the performance measures for all the NDT technologies are summarized in Table 7.7. Unlike the initial grading based on the literature search, the grading herein is made only for the deterioration types for which the technol- ogy was validated. For technologies that were represented by multiple participants, the performance measure grades are represented by mean values for the same group. Cer- tainly, an argument can be made that a technology should be represented by its best performers. In this case, the perfor- mance and corresponding grades would represent the true potential of the technology at this time. Also, the grading of the cost in the table is based on the cost, including the antici- pated traffic control costs. The overall grades span from 2.3 to 3.6, which puts all the technologies in the fair-to-good range with respect to the evaluation of a particular deterioration type. The overall good grades for a particular deterioration type by multiple technologies, however, come from different performance measures. The strengths of some technologies are higher accuracy and precision (repeatability), whereas some tech- nologies are stronger in other performance measures, such as speed or cost. For example, impact echo, GPR, infrared ther- mography, and chain dragging and hammer sounding receive close grades for delamination detection. However, impact echo’s higher grade derives primarily from the accuracy, GPR’s and infrared thermography’s grades from the speed, and chain dragging’s grade from the ease of use. Therefore, a decision regarding the selection of the technology should also be based on the most important performance measure. In the case of corrosion, the selection should be guided by the objective of the assessment. Four technologies received good grades for corrosion characterization: electrical resistivity, 27% 65% 384% 43% 205% $- $0.20 $0.40 $0.60 $0.80 $1.00 $1.20 $1.40 Radar Acousc Chain Dragging and Hammer Sounding Infrared Electrochemical O ve ra ll Co st / 2 Excluding Traffic Control Including Traffic Control Figure 7.5. The overall cost of survey (with and without traffic control). Table 7.6. Grading of NDT Technologies Based on Cost Radar Acoustic Chain Dragging and Hammer Sounding Infrared Thermography Electrochemical Grades (excluding traffic control) 4 3 5 5 5 Grades (including traffic control) 3 2 3 4 4

77 half-cell potential, galvanostatic pulse measurement, and ground-penetrating radar. However, ER and GPR will describe the corrosive environment of concrete to a greater extent, while GPM and HCP will describe corrosion rate and activity, respectively. The overall value of the NDT technology in concrete bridge deck deterioration detection is summarized in Table 7.8. The deterioration type grades were taken from Table 7.7, where the technology was applicable. Applications of IE in vertical crack depth and concrete quality estimation, and GPR in con- crete degradation assessment, were described in Chapter 2. Because those applications were not validated during the study, the technologies received a grade of 1 for the same. The technology that provides the highest value is GPR. The second most valuable technologies are impact echo and ultrasonic surface waves. However, the ultimate decision on which equip- ment to acquire and which technology to use will depend on a number of elements. Among others, it will depend on the type of deterioration that is of the highest concern to the agency and whether the evaluation is being done for network- level condition monitoring, or the project level for mainte- nance or rehabilitation. Table 7.7. Deterioration Type Grades for Validated NDT Technologies NDT Technology Deterioration Type Accuracy Precision (Repeatability) Speed Ease of Use Cost Overall Deterioration WF = 0.3 WF = 0.3 WF = 0.2 WF = 0.1 WF = 0.1 Type Grade Impact echo Delamination 2.8 4.0 2.3 2.1 3.0 3.0 Ultrasonic surface waves Delamination 2.8 3.0 2.4 1.4 3.0 2.7 Crack depth 2.5 3.0 1.0 1.4 3.0 2.3 Concrete deterioration 3.8 4.0 2.4 1.4 3.0 3.3 Ground-penetrating radar Delamination 2.1 4.0 3.9 2.2 3.0 3.1 Corrosion 1.6 4.0 3.9 2.2 3.0 3.0 Half-cell potential Corrosion 3.0 3.0 3.8 3.4 4.0 3.3 Galvanostatic pulse measurement Corrosion 2.4 3.0 2.4 2.6 4.0 2.8 Electrical resistivity Corrosion 3.0 4.0 3.8 3.6 4.0 3.6 Infrared thermography Delamination 2.2 2.0 4.1 4.0 4.0 2.9 Chain dragging Delamination 2.2 3.0 3.2 4.0 3.0 2.9 Note: WF = weight factor. Table 7.8. Overall Value of NDT Technology in Bridge Deck Deterioration Detection Deterioration Type Delamination Corrosion Vertical Cracks Concrete Degradation Overall Value RankingWF = 0.42 WF = 0.35 WF = 0.10 WF = 0.13 Impact echo 3.0 0.0 1.0 1.0 1.5 2 Ultrasonic surface waves 2.7 0.0 2.4 3.3 1.8 2 Ground-penetrating radar 3.1 3.1 0.0 1.0 2.5 1 Half-cell potential 0.0 3.3 0.0 0.0 1.2 3 Galvanostatic pulse measurement 0.0 2.8 0.0 0.0 1.0 3 Electrical resistivity 0.0 3.6 0.0 0.0 1.3 3 Infrared thermography 2.9 0.0 0.0 0.0 1.2 3 Chain dragging/hammer sounding 2.9 0.0 0.0 0.0 1.2 3 Note: WF = weight factor.