1、STP-NU-069ANALYSIS OF SELECTED NONDESTRUCTIVE EXAMINATION (NDE) METHODOLOGIES FOR THE ASSESSMENT OF CRACKING IN CONCRETE CONTAINMENTSSTP-NU-069 ANALYSIS OF SELECTED NONDESTRUCTIVE EXAMINATION (NDE) METHODOLOGIES FOR THE ASSESSMENT OF CRACKING IN CONCRETE CONTAINMENTS Prepared by: Maria Guimaraes Dav
2、id B. Scott Paul Weeks Randall Manley James Wall Electric Power Research Institute (EPRI) Date of Issuance: March 14, 2014 This report was prepared as an account of work sponsored by ASME Nuclear Codes so, delamination may be created using a natural occurring phenomenon such as corrosion. Delaminati
3、on was successfully created with this technique. However, scaling this up to larger mockups (needed for our tests) was not successful and therefore only the technique of embedding flaws was used for these tests. STP-NU-069: Analysis of Selected Nondestructive Examination (NDE) Methodologies for the
4、Assessment of Cracking in Concrete Containments 6 4 SHEAR WAVE TOMOGRAPHY The data obtained through shear wave tomography is usually displayed in C, B or D scans. The system output is user-friendly and easy to manipulate to obtain signals. Shear wave tomography can be used to detect defects at diffe
5、rent depths. 4.1 Depth of Defects Detection of delamination at approximate 0.30 m (12 in) depth. Shear wave tomography tests were performed at Crystal River with the objective of evaluating the capability of this equipment to detect a deep delamination. A shear wave tomographer with a 4 x 10 array o
6、f transducers was used. Figure 4-1 shows the B, C and D scans from a delaminated area. Note that a single layer of rebar was present between the surface and the delamination in this image, which largely facilitated the identification of the flaw. The area chosen for this demonstration was later show
7、n to have a crack width of about 2.5 cm (1 in). For more information on this particular case see 1. Detection of voids located at 0.30 m (12 in) below the surface. In this case, a smaller shear wave tomographer consisting of a 4 x 6 transducer array was used with the objective of determining the dep
8、th of voids previously embedded in the mockup shown in the top left corner of Figure 4-2. C-scans at different depths are also shown with the voids marked by a circle at depths of 20 cm and 31 cm. Note that despite the fact that several layers of reinforcement were present between the surface and th
9、e voids, the voids can be clearly seen at 0.20 m and 0.30 m (8 in and 12 in) deep. STP-NU-069: Analysis of Selected Nondestructive Examination (NDE) Methodologies for the Assessment of Cracking in Concrete Containments 7 Figure 4-1: Results from the Shear Wave Tomography Tests Performed at Crystal R
10、iver, showing the Delaminated Area Delamination Rebar Tested area STP-NU-069: Analysis of Selected Nondestructive Examination (NDE) Methodologies for the Assessment of Cracking in Concrete Containments 8 Figure 4-2: C-scans Slices at Depths of 7, 20, 31, and 37 cm Both voids could be located (marked
11、 with circles) d e p th 7 c md e p th 3 7 c md e p th 3 1 c md e p th 2 0 c mSTP-NU-069: Analysis of Selected Nondestructive Examination (NDE) Methodologies for the Assessment of Cracking in Concrete Containments 9 Detection of flaws located at 7.6 cm (3 in) and 15.2 cm (6 in) below the surface. Fig
12、ure 4-3 shows the B and C scans from the mockups referenced in Figure 3-1 showing delaminations located at 7.6 cm (3 in) and 15.2 cm (6 in) depth. Main observations follow: The depth of the back wall (25 cm / 10 in) can be easily identified. There is a blind zone near the surface of about 5-7 cm (2-
13、3 in), where defects cannot be seen by the ultrasonic shear waves. The defect positioned at 7.6 cm (3 in) depth is seen at 10-12 cm (5-6 in) depth. The defect positioned at 15.2 cm (6 in) depth is seen approximately at 20 cm (8 in) depth. The depth at which features are observed with this device nee
14、ds to be corrected to reflect the actual measurement (not done here). This method appears very robust to give an idea of the location of the defect, but precise depth determination does not seem accurate. Note that the reinforcing bars located near the back wall cannot be identified and get merged w
15、ith the images of the back wall. Figure 4-3: C-scans at Depths of 10 cm B scans taken over red dotted line 4.2 Operator Dependence In order to test the operator dependence, two sets of tests were performed by two different operators. Main observations follow: Bc a r d b o a r dp o ly e thp r e fa bC
16、p o ly e thBCSh a llo w d e fe c tD e e p e r d e fe c tSTP-NU-069: Analysis of Selected Nondestructive Examination (NDE) Methodologies for the Assessment of Cracking in Concrete Containments 10 There is little to no operator dependence. These tests were performed by two different operators. The res
17、ults are very consistent when comparing both data sets. Figure 4-4 shows an example from data at a location where one of the defects has a slight tilt. The data acquisition and data processing of this device is quite straight forward for simple geometries and in case of delaminations. However, in ca
18、ses of complex geometry, or if tendons or other materials (such as steel liners) are present, the signal processing is more involved and may require the use of software that is not provided by the commercially available device. In that case, the A-scans can be extracted and the data is processed thr
19、ough another system. Furthermore, phase evaluation of the signals may result in more refined images that allow differentiating empty voids with voids filled with water. In that case, the raw data needs to be extracted and analyzed through additional software. Figure 4-4: Comparison of Results from T
20、wo Operators Note: the technique does not seem to be operator dependent 4.3 Ease of Deployment Automation The speed of testing is slow when compared to the other tests performed here (impact echo and impulse response). In general, in order to obtain a good definition the tests are performed by overl
21、apping shots. This device has been previously mounted in a robot used to inspect horizontal surfaces like bridges or parking decks 2. Hence, automation is possible, but this device and the mode of testing ares rather slow for scanning large areas. This device is more suitable for getting in depth in
22、formation of a rather small area, than to scan large areas such as a containment structure. O p e r a to r 1O p e r a to r 2CBCBSTP-NU-069: Analysis of Selected Nondestructive Examination (NDE) Methodologies for the Assessment of Cracking in Concrete Containments 11 5 IMPACT ECHO The data obtained t
23、hrough impact echo is usually displayed in a frequency domain with main frequency peaks representing depth of delaminations or back wall. The system output is not very user friendly. Multiple reflections can be a source of error when using this technique. In this particular case, the edges of the mo
24、ckup generate reflections that make it very difficult to characterize the regions close to the edges. Boundary effects and multiple reflections need to be accounted for when using this technique. 5.1 Depth of Defects This study shows detection capabilities of delaminations at 7.5 cm (3 in) and 15 cm
25、 (6 in) depth. Figure 5-1 shows the frequency signals from a section of the mockup covering both depths of defects. Main observations follow: Areas shaded in grey represent the part of the signals that are influenced by the boundary effects. Hence, the analysis is performed in the non-shaded areas.
26、The depth of the back wall cannot be determined here since it falls under the grey shaded area (boundary effects). As previously explained, the frequency if the main peaks determines the depth of the delamination or the back wall. In this case, the frequencies of concern are: 26 kHz for the shallow
27、flaw (7.5 cm) and 13 kHz for the deeper flaw (15 cm). Impact echo plots can be presented in different forms. Here, the signals in the frequency domain are shown in the form of a cascade graph, with one signal per point tested. This study shows delaminations up to a depth of 15 cm (6 in). In general,
28、 the majority of studies on impact echo are focused on transportation infrastructure and the depths of delamination are shallow (usually less than 15 cm). However, this technique can be used for deeper defects as long as the problem with multiple reflections is overcome. Wiggenhauser 3 shows tests o
29、n concrete structures with a back wall of 83 cm (33 in), and reports that the depth of the back wall could be reliably identified. STP-NU-069: Analysis of Selected Nondestructive Examination (NDE) Methodologies for the Assessment of Cracking in Concrete Containments 12 Figure 5-1: Impact Echo Signal
30、s Shown Here in Frequency Domain Note: shaded areas denote boundary affected area; red lines denote characteristic frequencies for each depth; the lower right corner shows the area of the mockup tested 5.2 Operator Dependence Impact echo is the technique among the ones tested here that is more opera
31、tor dependent and requires higher level of training. Signal interpretation is not straight forward and repeatability and consistency are not guaranteed. Furthermore, multiple reflections due to boundary effects, honeycombs, non-planar defects, etc., complicate the signal interpretation. Operator tra
32、ining is crucial and absolutely required for data collection and analysis with this technique. A large amount of research is being performed on the interpretation of these tests, which has the potential to overcome some of the limitations with signal interpretation 3. In nuclear plants ultrasonic te
33、chniques are commonly used to perform examinations in primary and secondary systems. The need of having trained and certified inspectors to perform inspections in safety related components has been recognized for some time in the nuclear industry. Training requirements, certification programs and qu
34、alification of procedures and personnel are required by the ASME code for ultrasonic in-service inspection of safety related piping and components in nuclear power plants. 3”6”STP-NU-069: Analysis of Selected Nondestructive Examination (NDE) Methodologies for the Assessment of Cracking in Concrete C
35、ontainments 13 There are no certifications or training available to perform NDE tests on concrete structures today. Hence, when a plant is required to perform one of these tests it relies on vendor expertise with little verifiable qualifications. The American Concrete Institute (ACI) is developing a
36、 training certification for NDE devices, which is a good starting point to address this challenge. However, certification should not be confined only to training, but also should require a set amount of hours of field testing, conducing to a similar program as the one for Level II and Level III insp
37、ectors. 5.3 Ease of Deployment Automation Impact echo devices have the potential for deployment in automated testing of large structures. Vertical concrete structures, such as containments, could be inspected with this method as it is being explored by EPRI through its concrete strategic program ini
38、tiative 45. Note that impact echo devices have previously been mounted on small robots for inspection of bridge decks and parking decks, generally on horizontal surfaces 2. Testing and data collection can be performed rather fast covering large areas in short time. A relatively new development on im
39、pact echo testing is the incorporation of non-contact receivers to capture the signals, normally called “air-coupled impact echo” 6. Air-coupled impact echo is not yet a commercially available technique, but it is an R blue: intact) and a picture of the locations where cores were take (red dots: del
40、aminated core; yellow dot: clean core). Note that although the delamination was deep, the cracks were rather wide making it easier to detect the flexural mobility of the delaminated area. The extensive tests performed at Crystal River have shown the adequacy of this technique for that type of delami
41、nation. 0.10 m to 0.38 m (4 to 15 in) deep cracks with widths of less than 0.03 cm (0.013 in). The cracks encountered at Davis-Besse shield building were thinner and shallower than in Crystal River. This technique was the preferred NDE method for determining the extent of condition after cracks were
42、 identified 8. Again, the technique performed very well and allowed for a 100% testing of the shield wall as shown in tests in Figure 6-2. Shallower cracks can also be identified with impulse response as shown in numerous publications, as summarized in 9. When a wall is too thick, or the crack too d
43、eep, the delaminated structure does not behave like a plate and the reliability of this technique may be compromised. STP-NU-069: Analysis of Selected Nondestructive Examination (NDE) Methodologies for the Assessment of Cracking in Concrete Containments 15 Figure 6-1: Impulse Response Tests in Cryst
44、al River Containment Wall Note: Shaded blue shows non-delaminated area; yellow shows transition zone; shaded red (left) or white (right) shows delaminated area Left hand side shows the accuracy of the impulse response readings, with red dots representing cores taken showing delamination and green do
45、ts representing intact cores. Right hand side shows the widths of the cracks within the hour glass delaminated shape 7 STP-NU-069: Analysis of Selected Nondestructive Examination (NDE) Methodologies for the Assessment of Cracking in Concrete Containments 16 Figure 6-2: Impulse Response Tests in Davi
46、s Besse Shield Building Note: Grey shows non-delaminated area; yellow shows transition zones; purple shows delaminated areas 8 6.2 Operator Dependence Impulse response is a reliable technique. However, operator training is required, albeit it is less necessary than with impact echo. 6.3 Ease of Depl
47、oyment Automation Testing and data collection are performed rather fast compared with other techniques like ultrasound imaging or impact echo. Hence, large areas can be covered in a relatively short time. Impulse response would be an excellent method for application in an automated device. However,
48、the main concern is the size and weight of the impactor, which will be challenging to automate, though not impossible. STP-NU-069: Analysis of Selected Nondestructive Examination (NDE) Methodologies for the Assessment of Cracking in Concrete Containments 17 7 DEPLOYMENT: AUTOMATION OF NDE INSPECTION
49、S OF CONCRETE STRUCTURES The infrastructure in the energy industry includes several types of large, curved vertical structures such as cooling towers, nuclear containments and hydroelectric dams. Inspections of these structures are often performed manually by means of scaffolds. Automating such inspections will reduce time and costs, make them more efficient, increase inspection frequency and reduce safety risks. The first step in automating large civil infrastructure inspections was to search on other industries
