National Academies Press: OpenBook

Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models (2013)

Chapter: Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves

« Previous: Chapter 2 - Theoretical Models for Infrared Thermography Technology
Page 12
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 12
Page 13
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 13
Page 14
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 14
Page 15
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 15
Page 16
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 16
Page 17
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 17
Page 18
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 18
Page 19
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 19
Page 20
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 20
Page 21
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 21
Page 22
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 22
Page 23
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 23
Page 24
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 24
Page 25
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 25
Page 26
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 26
Page 27
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 27
Page 28
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 28
Page 29
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 29
Page 30
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 30
Page 31
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 31
Page 32
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 32
Page 33
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 33
Page 34
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 34
Page 35
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 35
Page 36
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 36
Page 37
Suggested Citation:"Chapter 3 - Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves." National Academies of Sciences, Engineering, and Medicine. 2013. Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models. Washington, DC: The National Academies Press. doi: 10.17226/22603.
×
Page 37

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.

12 Theoretical Models for Mechanical Wave Technology: Impact Echo, Impulse Response, and Ultrasonic Surface Waves The numerical simulations presented in this chapter were carried out by Infrasense with support from Dr. Kim Belli of Northeastern University in Boston, Professor Dennis Hiltunen of the University of Florida, and Professor Rajib Mallick of Worcester Polytechnic Institute. Numerical simulations of nondestructive testing (NDT) techniques were carried out to assess the NDT ability to detect delaminations and to evaluate the most promising configu- rations for implementing each NDT method. An NDT tech- nique and/or configuration that does not show promise in the numerical simulations would not likely succeed in the field evaluations; therefore, results obtained from the simulations would help the project team focus on the most promising methods or configurations, or both, in the laboratory and field evaluations. The numerical simulations were conducted by using existing numerical models to simulate and compare simu- lation results of delaminated and intact pavements. The simulation information was used to define detectability and served to support recommendations of specific meth- ods or configurations, or both, for the laboratory and field evaluations. Modeling provided insight into results that could be expected from laboratory and field experiments, and also offered guidance into the effective configurations of the transmitter and receiver for each NDT technique. Modeling results were considerably cleaner than were results obtained in the laboratory and field evaluations, and mod- eling results were based on certain idealizations of material characteristics and input signals. Simulation results showed (a) whether, under ideal conditions, the pavement defect could be detected; and (b) under what conditions detectabil- ity could be enhanced. For electromagnetic and mechanical wave methods, the simulation results could show how the relative placement of the source and receiver could affect the NDT detectability. Introduction This study explores the possibilities of using mechanical waves to detect flexible pavement with delamination. When finite element software was used, the behavior of sound pave- ment was evaluated and a parametric study on delaminated pavement was then elaborated. By comparing the differences in dispersion images, the parametric study shows how the impacts of different variables are related to the properties of the pavement with delamination. Previous Work Munoz (2009) conducted an extensive evaluation of capa- bilities of mechanical wave testing methods to characterize pavements containing delaminations. The evaluation used finite element simulations, and included impact echo (IE), impulse response, and ultrasonic surface waves (USW) meth- odologies. An overview of this study is provided in the following sections. Finite Element Model Important features of the finite element models used in the work were as follows: • The finite element analysis was conducted with the commercial software LS-DYNA 3D. • For USW and IE methodologies, the pavement model was 0.5 × 0.5 × 0.5 m in size and used 10 × 10 × 10 mm 8-node solid elements. • For the impulse response methodology, the pavement model was 2 × 2 × 0.5 m in size and used 25 × 25 × 25 mm 8-node solid elements. • The pavement models contained seven layers: four 50-mm layers for the hot-mix asphalt (HMA), two 100-mm layers C h a P t E r 3

13 for the base, and one 100-mm layer for the subgrade, for a total depth of 0.5 m. • Nonreflecting boundaries were used to absorb energy. • An impact source was modeled with a half-sine curve with duration of 52.5 µs for the USW and IE methodologies, and 0.2 ms for the impulse response methodology. Parameters Six parameters of the pavement models were investigated with the finite element simulations. The parameters were • Degree of defect = bonded, partially debonded, and totally debonded interface conditions; • Depth of defect = 50, 100, and 150 mm from surface; • Size of defect = 100 × 100 mm, 300 × 300 mm, and 500 × 500 mm in horizontal extent; • HMA modulus = 12.5, 8.3, 6.25, and 4.2 GPa; • Base modulus = 1,250, 700, and 315 MPa; and • HMA thickness = 200, 150, and 100 mm. Impact Echo The IE methodology was investigated by using a standard one- impact source and one-receiver configuration. The significant findings were as follows: • Degree of defect: totally and partially debonded defects were detectable and accurately located in frequency domain (Figure 3.1). • Defect depth: totally debonded defects were detectable and accurately located in frequency domain for depths of 100 and 150 mm; for 50 mm, the method generated a dominant flexural mode (Figure 3.2). • Defect size: totally debonded defects were detectable and accurately located in frequency domain for 300 × 300 mm and 500 × 500 mm defects; for 100 × 100 mm, the dominant frequency was near the thickness of HMA (some diffraction effect due to defect), while the defect was correctly located via a smaller peak (Figure 3.3). • HMA modulus: totally debonded defects were detectable and accurately located in frequency domain, but the HMA modulus needed to be known for accurately locating the defect (Figure 3.4). • Base modulus: totally debonded defects were detectable and accurately located in frequency domain independent of base modulus (Figure 3.5). • HMA thickness: totally debonded defects were detect- able and accurately located in frequency domain inde- pendent of HMA thickness for a given defect location (Figure 3.6). Figure 3.1. Degree of defect: (a) time record signals, (b) time record after applying the Blackman-Harris window, and (c) frequency spectra of signals. Source: Munoz 2009. (a) (b) (c)

14 Source: Munoz 2009. (a) (b) (c) Figure 3.2. Defect depth: (a) time record signals, (b) time record after applying the Blackman-Harris window, and (c) frequency spectra of signals. Source: Munoz 2009. (a) (b) (c) Figure 3.3. Defect size: (a) time record signals, (b) time record after applying the Blackman-Harris window, and (c) frequency spectra of signals.

15 Source: Munoz 2009. (a) (b) (c) Figure 3.4. HMA modulus: (a) time record signals, (b) time record after applying the Blackman-Harris window, and (c) frequency spectra of signals. Source: Munoz 2009. (a) (b) (c) Figure 3.5. Base modulus: (a) time record signals, (b) time record after applying the Blackman-Harris window, and (c) frequency spectra of signals.

16 Impulse Response The impulse response methodology was investigated by using a standard one-impact source and one-receiver configuration. The significant findings were as follows: • Degree of defect: totally and partially debonded defects could be differentiated from the intact response (Figure 3.7). • Defect depth: totally debonded defects could be differ- entiated from intact response for depths of 50 and 100 mm; for 150 mm, the response with defect was similar to intact response (Figure 3.8). • Defect size: totally debonded defects could be differentiated from intact response for defect sizes of 500 × 500 mm; for 300 × 300 mm and 100 × 100 mm, the responses with defect were similar to intact response (Figure 3.9). • HMA modulus: totally debonded defects could be differ- entiated from intact response, but response was depen- dent on HMA modulus (Figure 3.10). • Base modulus: totally debonded defect response was independent of base modulus (Figure 3.11). Figure 3.7. Degree of defect: (a) signal from geophone and (b) Fast Fourier Transform (FFT) hammer/geophone. Source: Munoz 2009. (a) (b) Figure 3.6. HMA thickness: (a) time record signals, (b) time record after applying the Blackman-Harris window, and (c) frequency spectra of signals. Source: Munoz 2009. (a) (b) (c)

17 Figure 3.8. Defect depth: (a) signal from geophone and (b) FFT hammer/geophone. Source: Munoz 2009. (a) (b) Figure 3.9. Defect size: (a) signal from geophone and (b) FFT hammer/geophone. Source: Munoz 2009. (a) (b) Figure 3.10. HMA modulus: (a) signal from geophone and (b) FFT hammer/geophone. Source: Munoz 2009. (a) (b)

18 • HMA thickness: totally debonded defect responses were similar for thicknesses of 150 and 200 mm; the response was smaller for 100 mm (Figure 3.12). Ultrasonic Surface Waves The USW methodology was investigated by using a standard configuration of one impact source and two receivers, and was intended to simulate a PSPA test device. The significant findings were as follows: • Degree of defect: totally debonded defects could be dif- ferentiated from intact response; the response was simi- lar to intact for partially debonded defects (Figure 3.13). • Defect depth: totally debonded defects could be differen- tiated from intact response for depths of 50 and 100 mm; the response was similar to intact response for 150 mm (Figure 3.14). • Defect size: totally debonded defects could be differentiated from intact response for defect sizes of 500 × 500 mm and 300 × 300 mm; the response was similar to intact response for 100 × 100 mm (Figure 3.15). • HMA modulus: totally debonded defects could be differ- entiated from intact response, but response was depen- dent on HMA modulus (Figure 3.16). • Base modulus: the response from totally debonded defects was independent of base modulus (Figure 3.17). • HMA thickness: the response for totally debonded defects was similar for thicknesses of 150 and 200 mm, but slightly different for a thickness of 100 mm (Figure 3.18). Figure 3.11. Base modulus: (a) signal from geophone and (b) FFT hammer/geophone. Source: Munoz 2009. (a) (b) Figure 3.12. HMA thickness: (a) signal from geophone and (b) FFT hammer/geophone. Source: Munoz 2009. (a) (b) (text continued on page 25)

19 Figure 3.13. Degree of defect: (a) dispersion curves and (b) average velocity of surface waves. Source: Munoz 2009. (a) (b)

20 Source: Munoz 2009. (a) (b) Figure 3.14. Defect depth: (a) dispersion curves and (b) average velocity of surface waves.

21 Source: Munoz 2009. (a) (b) Figure 3.15. Defect size: (a) dispersion curves and (b) average velocity of surface waves.

22 Source: Munoz 2009. (a) (b) Figure 3.16. HMA modulus: (a) dispersion curves and (b) average velocity of surface waves.

23 Source: Munoz 2009. (a) (b) Figure 3.17. Base modulus: (a) dispersion curves and (b) average velocity of surface waves.

24 Source: Munoz 2009. (a) (b) Figure 3.18. HMA thickness: (a) dispersion curves and (b) average velocity of surface waves.

25 Numerical Modeling As reviewed in the previous sections, Munoz (2009) has demonstrated the significant capability of mechanical wave methods to characterize pavements with delaminations. One of the serious limitations revealed in the study was the inabil- ity of PSPA-based (two-sensor) ultrasonic surface waves to distinguish partial debonding. Therefore, further finite ele- ment simulations were conducted to investigate whether multiple-sensor surface wave techniques could provide better characterization of defects. Multiple-sensor techniques have been widely implemented in geotechnical site investigations in recent years because of the improved resolution that can be achieved. The multiple-sensor techniques also provide a more fundamental understanding of the characteristics of the wave field, and it was anticipated that this advanced understanding could further resolve delamination defects in pavements. Delamination is generally known as a weak bonding between adjacent HMA layers, producing a sliding effect between layers and thus creating flaws such as slippage and potholes at later stages. In this study, the delaminations were modeled with two approaches: (1) a very thin layer (2-mm thick) with low modulus and (2) an interface element introduced between HMA lifts. For the numerical study, a number of criteria have been taken into account to ensure accuracy of the results. The criteria con- sidered were the dimensions of model, element size, sampling rate, impact frequency, and total time of recording, as follows: • Axisymmetric models were used to simulate three- dimensional (3-D) wave propagation in reality, with dimensions 2 m wide by 1 m deep (Figure 3.19). • 15-node triangular elements were used with an effective element length of 1 cm. • The energy was created via a vertical impact source mod- eled as a triangular time pulse with duration of either 100 µs (10 kHz) or 50 µs (20 kHz), and was located at the upper left corner of the model (Figure 3.19). • The output time histories from the PLAXIS model were in terms of vertical particle velocities collected along the top HMA surface via a linear, nonuniform array at 20 locations (Figure 3.20). This array was intended to simulate real data collection in a multichannel analysis of surface waves (MASW) test by using geophone velocity transducers on a pavement specimen. As done in the field, the array has been optimized to produce a high-quality dispersion image. The dense sensors at the beginning were to ensure high frequency resolution, while the array length of 0.6 m was to ensure that surface waves with long wavelengths (low frequencies) were included in the collected signals. Also note that the initial 0.5-m offset of the source from the first receiver was to reduce the so-called near-field effect at low frequencies. • The sampling rate for the output time histories was chosen to be 0.1 µs, and the total record length was 10 ms. • As shown in Figure 3.19, the model was divided into sev- eral different layers from top to bottom: three layers with a thickness of 5 cm each for the HMA layers, one layer with a thickness of 25 cm to represent the base, and one layer with a thickness of 0.60 m for the subgrade, for a total depth of 1 m. • As indicated in Figure 3.19 via the PLAXIS symbols, the left and right boundaries were fixed from horizontal trans- lation but were allowed to move vertically, while the bottom boundary was fixed from both horizontal and vertical translation. Figure 3.19. Pavement model layout.

26 • Absorbing boundaries were applied along the bottom and right edges to simulate the continuity of the materials, thus eliminating reflecting waves bouncing back from the boundaries. Parametric Study The parametric study was conducted by using a control model as described above and having material properties as indicated in Table 3.1. The effects of the parameters shown in Table 3.2 were then systematically investigated, as follows: • Three defect depths of 5, 10, and 15 cm were investigated for an HMA layer with Vs = 888 m/s (control). • Three HMA stiffnesses having shear wave velocities of Vs = 400, Vs = 888, and Vs = 1,250 m/s were investigated for a defect at 5 cm depth. • The delamination was modeled in two ways: a soft, thin layer and an interface element; the severity of the delamination was investigated with HMA Vs = 888 m/s and a defect at 5 cm depth. 4 Thin layer with low modulus (modulus of air, modulus of base); and 4 Interface element with R factor of 0.01 and 0.22. • The size (horizontal extent) of the delamination was inves- tigated with HMA Vs = 888 m/s and a defect at 5 cm depth. 4 The test array directly above the delamination; 4 The test array partially in front of the delamination; and 4 The test array partially after the delamination. results Control Model The presented results are so-called dispersion images based on a multichannel analysis of all types of seismic waves propa- gating along the surface of a pavement. The dispersion images Length (m) W id th (m ) Figure 3.20. Nonuniform array. Table 3.1. Properties for Pavement Profile—Control Model Material Unit Weight (kN/m3) Poisson’s Ratio Vp (m/s) Vs (m/s) Layer Thickness (m) HMA 23 0.35 1,850 888 0.15 Base 19 0.35 407 195 0.25 Subgrade 15 0.35 324 156 0.6 Note: Vp = phase velocity. Table 3.2. Parameter Study for Evaluating Pavement with Delaminations Variable Description Depth of defect At 5, 10, and 15 cm Modulus of HMA With Vs = 400, 888, and 1,250 m/s Bonding condition Thin layer filled with air or base Interface with strength reduction R = 0.01 and 0.22 Size of defect Full extent or partial coverage of test array

27 are a form of 3-D power spectrum in which the surface wave phase velocity versus frequency (dispersion) of the pro- pagating waves is displayed on the horizontal axes, and the energy present at each velocity-frequency pair is displayed via a color coding, with the cold colors corresponding to low energy, and the hot colors corresponding to high energy. A concentration of hot energy over a narrow band represents a normal mode of wave propagation, and the velocity-frequency pairs along the peak of the narrow band are typically referred to as the dispersion curve in surface wave testing. A dis persion curve is a relationship showing how the surface wave phase Figure 3.22. Dispersion image of intact pavement with Vs 5 400 m/s for HMA with pulse frequency of 10 kHz. velocity of a layered material changes with frequency (or wavelength). In this study, the dispersion images are obtained via a cylindrical beamforming algorithm, which essentially transforms the raw signals (i.e., the time signatures obtained from PLAXIS) from the time-space domain to the velocity- frequency domain, by taking into account the cylindrically propagated wave field. In this study, intact (no defect) pavement models were treated as control models for comparison with various defect models. It was found that the dispersion images from the intact pavement profiles (Figures 3.21 to 3.23) did not display Figure 3.21. Dispersion images of intact pavement with Vs 5 888 m/s for HMA (a) with pulse frequency of 20 kHz and (b) with pulse frequency of 10 kHz. (a) (b) Figure 3.23. Dispersion image of intact pavement with Vs 5 1,250 m/s for HMA with pulse frequency of 20 kHz.

28 image (a) and image (b) in Figure 3.21 are nearly the same. For this case of total debonding, the HMA layer above can freely slide along the interface, and this layer can essentially be thought of as a free plate. Hence, the results can be explained by Lamb’s free plate model as shown in Figure 3.27. Lamb waves refer to those in thin plates (with planar dimensions being far greater than that of the thickness, and with the wavelength being on the order of the thickness) that provide upper and lower boundaries to guide continuous propaga- tion of the waves. Lamb waves consist of symmetric (Si) and antisymmetric (Ai) modes, which can coexist when excited. The antisymmetric modes correspond to bending waves in the plate, and the symmetric modes correspond to quasi- longitudinal waves in the plate. Figure 3.28a plots the Lamb wave dispersion curves in a free plate. At high frequencies, the A0 and S0 modes merge together, and the convergent velocity corresponds to the Rayleigh wave velocity of the free plate. For example, in Figure 3.24, it is clear that the dominating mode is the fundamental antisymmetric mode (A0) of Lamb waves. In addition, there is another branch trying to merge with A0 from above, which is expected to be the fundamental symmetric mode (S0). Second, on comparing the intact model from Figure 3.21 with the defect model in Figure 3.24, another important feature is revealed: at low frequencies the dispersion curve (narrow band) of the delaminated case appears to be much smoother, and this smooth curve is interpreted as the A0 mode. The dispersion image of the intact model has the discontinuities associated with the base and subgrade inter- action, but the defect model is free of these low-frequency effects due to the debonding. For the cases of delamination at 10 cm (Figure 3.25), the dispersion image shows discontinuities at approximately 6 kHz and 11 kHz. The wavelengths are to (a) (b) Figure 3.24. Dispersion images of total delamination at a depth of 5 cm for (a) air and (b) R 5 0.01. a continuous band of normal mode energy, but displayed several branches, especially at low frequencies. It is believed that such discontinuities correspond with interaction between the high-modulus surface layer and the base and subgrade in the pavement system (Rydén 2004). However, at high frequencies, a narrow band (dispersion curve) tends to con- verge to a single value, which corresponds to the Rayleigh wave velocity of the very top HMA layer (slightly lower than Vs and according to Poisson’s ratio). Depth of Defect To study the effect of a defect, total debonding is introduced at depths of 5, 10, and 15 cm, respectively, across the full horizontal width of the model, and is intended to represent complete shallow, intermediate, and deep delaminations in reality. Note that for the cases of 5 cm and 10 cm, the defect is in the HMA layer; for the case of 15 cm, the defect is right at the interface between HMA and base. Total debonding is modeled as either an extremely thin layer filled with air, or via an interface element with the strength reduction factor set to its minimum (R = 0.01). The strength reduction factor (R) is a parameter employed in PLAXIS to control the degree of continuity between two adjacent surfaces, having a maximum value of 1 (complete continuity) and a minimum value of 0.01 (nearly independent or frictionless). To illustrate the model, a HMA shear wave velocity of 888 m/s is considered, and the dispersion images for the defect models are shown in Figures 3.24 to 3.26. Note that the corresponding dispersion image for the intact model has been shown in Figure 3.21. First, it should be observed that the two approaches to modeling the defect result in very simi- lar dispersion images at all defect depths considered; that is,

29 (a) (b) Figure 3.25. Dispersion images of total delamination at a depth of 10 cm for (a) air and (b) R 5 0.01. (a) (b) Figure 3.26. Dispersion images of total delamination at a depth of 15 cm for (a) air and (b) R 5 0.01. Source: Rydén 2004. Figure 3.27. Lamb’s free plate model.

30 some extent related to the depth of the delamination. For the cases of delamination at 15 cm (Figure 3.26), the dispersion image shows discontinuities at approximately 4 kHz and 6 kHz. It is observed that the discontinuities are more evi- dent than in the previous cases; that is, as the delamination goes deeper, the S0 mode becomes clearer. Alternatively stated, the dispersion image becomes similar to the intact case as the delamination goes deeper; thus, the surface wave method may not be very sensitive to deep delaminations. hMa Modulus Next, the effect of elastic modulus of the HMA layer is inves- tigated. It has been demonstrated that both approaches to modeling delamination yield very similar results. Considered here are only the cases with total debonding at a depth of 5 cm and modeled with a thin layer filled with air. Elastic moduli of the HMA layer are changed such that the shear wave velocities are 400, 888, and 1,250 m/s, accordingly. Figure 3.28 shows clearly that the dispersion curve is dominated by both the A0 and S0 modes of Lamb waves. Examining Figures 3.29 to 3.31 reveals that as the modulus of HMA increases, the S0 mode becomes weaker. In Fig- ure 3.31, the dispersion image is almost dominated by the A0 mode alone. It is also interesting to notice that for dif- ferent moduli of HMA, the fundamental symmetric mode (S0) appears at a higher starting frequency. For example, S0 appears from about 5.5 kHz for the HMA in Figure 3.30, and vaguely from about 17.5 kHz for the stiffest HMA in Figure 3.31. This result makes sense because Lamb wave Figure 3.28. (a) Lamb wave dispersion curves in free plate and (b) particle motion showing different types of waves in the model. Source: Rydén 2004. (a) (b) Lamb waves at Poisson’s ratio=0.35 2 1.8 1.6 1.4 1.2 1 0.8 0.6Ph as e ve lo ci ty /s he ar w av e v e lo ci ty 0.4 0.2 0 0 1000 2000 BENDING WAVE (ANTI-SYMMETRIC) QUASI-LONGITUDINAL WAVE (SYMMETRIC) RAYLEIGH WAVE 3000 4000 5000 6000 Bending wave velocity Frequency · thickness (Hzm) S0 A0 A1 S1 A2 S2 Quasi-longitudinal wave velocity Shear wave velocity Rayleigh wave velocity 7000 8000

31 Figure 3.29. Dispersion image of total delamination at a depth of 5 cm with HMA Vs 5 400 m/s. Figure 3.30. Dispersion image of total delamination at a depth of 5 cm with HMA Vs 5 888 m/s. dispersion is a function of plate modulus, as well as of plate thickness and Poisson’s ratio. Degree of Defect Total debonding between HMA lifts is probably rare in real- ity, and most often delaminations are characterized by weak bonds between interfaces. Therefore, partial delaminations are investigated by modeling the thin layer with a low modu- lus (base material) that is higher than air. To test whether the Figure 3.31. Dispersion image of total delamination at a depth of 5 cm with HMA Vs 5 1,250 m/s. Figure 3.32. Dispersion image of total delamination for thin layer filled with air at a depth of 5 cm and HMA Vs 5 888 m/s. interface approach would render similar results, the strength reduction factor is set to 0.22, which produces results com- parable to the base-filled thin layer. The results are presented for HMA Vs = 888 m/s and the defect at a depth of 5 cm. First, for aid in the comparison, Figure 3.32 is a repeated version of Figure 3.24, and Figure 3.33 displays the dis- persion image for a thin layer with low modulus higher than air. After comparing Figure 3.33 to Figure 3.32, it is evident that with the partial delamination the S0 mode nearly disappears. More over, Figure 3.33 shows a discontinuous low-frequency

32 response and a slight drop in phase velocity at high frequencies. It is known that the low-frequency discontinuity is the feature of an intact pavement profile. It is inferred that the slight drop in phase velocity at high frequencies may be related to the base-filled defect. Second, a very similar comparison is presented in Fig- ures 3.34 and 3.35 for the interface approach, including the lack of the S0 mode and a low-frequency discontinuity. Further more, it should be noted that the discontinuous low- frequency responses in Figures 3.33 and 3.35 are slightly dif- ferent from that in Figure 3.21, which implies that a shallow partial delamination gives rise to differences in the dispersion image at both high and low frequencies. Defect Size It is likely that delaminations are typically localized in size, and thus do not extend horizontally across the entire interface. Given a delaminated pavement specimen, the research team does not know where the delamination has occurred. Thus, if a test array is placed on the surface, the test array can by chance fully cover the size of the defect, or may just partially cover the delamination. Therefore, to simulate a real MASW test on pavement, three possible situations are considered: (1) an array right above the defect, (2) an array partially over the defect with the source and part of the array in front, and (3) an array with the source and part of the array over the defect. For full coverage, Case 1, the delamination length is set to be 1 m long horizontally, from 0.3 m to 1.3 m in the model (Fig- ure 3.19). For partial coverage, the delamination size is set to be 0.7 m long. The delamination extends from 0.85 to 1.55 m for Case 2 and from 0 to 0.7 m for Case 3. For all cases, the receiver array configuration is as shown in Figure 3.20, from 0.5 m to 1.1 m, and with the source at 0. The results are presented for HMA Vs = 888 m/s, and the defect at a depth of 5 cm. First, for aid in the comparison, Figure 3.36 is a repeated ver- sion of Figures 3.24b and 3.34: total debonding modeled with an interface element. For Case 1, in comparison with Figure 3.36, the dispersion image shown in Figure 3.37 contains discontinu- ities at low frequencies and no S0 mode. This can be explained from the deformed mesh at the end-of-time steps shown in Figure 3.41. Because the delamination starts from 0.3 m and Figure 3.33. Dispersion image of partial delamination for thin layer filled with base material at a depth of 5 cm. Figure 3.34. Dispersion image of total delamination for interface element with R 5 0.01. Figure 3.35. Dispersion image of partial delamination for interface element with R 5 0.22.

33 ends at 1.3 m, it is constrained from both ends. Therefore, a Lamb wave approximation is not appropriate in this case. For Case 2 shown in Figure 3.38, the dispersion image appears similar to the intact case previously presented in Figure 3.21, with some differences noted at low frequencies. Due to the fact that most of the sensors collect signals from an intact profile from 0.5 m to 0.85 m, the dispersion image at high frequencies is nearly the same as that in Figure 3.21. Thus, it is inferred that the slight low-frequency difference is created by sensors that are placed above the delamination. For Case 3 shown in Figure 3.39, the dispersion image gen- erally looks similar to the total delaminated case in Figure 3.36, with a noticeable S0 mode at high frequencies. However, it also displays a discontinuous A0 mode at low frequencies. In addi- tion to comparing dispersion images, it is helpful to also review the deformed meshes. Figure 3.40 plots the deformed mesh for the fully delaminated case corresponding to the disper- sion curve in Figure 3.36. With a full interface extension for the delamination, the flexural mode of vibration dominates. However, in Figures 3.41 and 3.42, the deformation is more constrained due to the confinement from the boundaries. In Figure 3.43, again the flexural mode dominates, and it happens right above the delamination. In summary, the inter pretation Figure 3.36. Dispersion image of total delamination with full interface extent. Figure 3.37. Dispersion image of total delamination with full array coverage (Case 1). Figure 3.38. Dispersion image of total delamination with partial array coverage (Case 2). Figure 3.39. Dispersion image of total delamination with partial array coverage (Case 3). (text continued on page 37)

34 Figure 3.40. Deformed mesh of total delamination with full interface extension.

35 Figure 3.41. Deformed mesh of partial delamination with full array coverage (Case 1).

36 Figure 3.42. Deformed mesh of partial delamination with partial array coverage (Case 2).

37 of dispersion images significantly depends both on the size of the delamination and the array configuration. Conclusions A surface wave technique has been applied to detect pave- ments with delamination. All results are based on finite ele- ment simulation and presented in the form of dispersion images. On the basis of differences in the dispersion images, the following conclusions were reached: 1. Delamination in HMA pavements can be effectively sim- ulated as either a thin layer with low modulus or by the interface element between layers. 2. In general, the shallow full-scale delamination is easier to detect and can be explained by the Lamb wave approxi- mation. The deep and small-scale delaminations are more difficult to detect. 3. Array configuration plays a significant role in interpret- ing differences in dispersion images caused by localized delaminations. References Munoz, D. 2009. Finite Element Modeling of Nondestructive Test Methods Used for Detection of Delamination in Hot Mix Asphalt Pavements. MS thesis, University of Texas at El Paso. Rydén, N. 2004. Surface Wave Testing on Pavements. PhD dissertation, Lund University, Lund, Sweden. (continued from page 33) Figure 3.43. Deformed mesh of partial delamination with partial array coverage (Case 3).

Next: Chapter 4 - Theoretical Models for Mechanical Wave Technology: Deflection-Based Approach »
Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models Get This Book
×
 Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 - Theoretical Models
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s second Strategic Highway Research Program (SHRP 2) Report S2-R06D-RW-2: Nondestructive Testing to Identify Delaminations Between HMA Layers, Volume 2 describes the theoretical models used in the development of nondestructive testing (NDT) techniques capable of detecting and quantifying delaminations in HMA pavements.

SHRP 2 Report S2-R06D-RW-2 was developed as part of SHRP 2 Renewal Project R06D, which generated a sizable amount of documentation regarding the findings of evaluations and equipment development. The report for SHRP 2 Renewal Project R06D is therefore divided into five volumes. Volume 1 is a comprehensive summary of the study. Volumes 2 through 5 provide more detailed technical information and are web-only. The topics covered in other volumes are listed below.

Volume 3: Controlled Evaluation Reports

Volume 4: Uncontrolled Evaluation Reports

Volume 5: Field Core Verification

Renewal Project R06D also produced a Phase 3 Report to document guidelines for use of ground penetrating radar and mechanical wave nondestructive technologies to detect delamination between asphalt pavement layers.

READ FREE ONLINE

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!