Grouting Defect Detection of Lapped Bar Connections Based on Impact-Echo Method

Grouted lap-splice connections are widely used for connecting precast concrete components. Grouting defects in the connections significantly influence the structural performance of the whole connection, which leads to the need for grouting defect detection. In this study, the impact-echo (IE) method was used for detecting defects in grouted lap-splice connections. Grouted connections with different levels of artificial grout defects were prepared in a shear wall, and the IE method was used to measure the frequency response. In addition, finite element (FE) analysis based on ABAQUS was conducted to simulate the tests. Based on the validated FEmodel, a parametric study was conducted to investigate the effect of the depth of the grout hole on the amplitude spectrum./e results indicated that (1) the IEmethod offered a good potential for grouting defect detection in grouted lap-splice connections; (2) the proposed FE model could well predict the frequency response of the grouting hole; and (3) the measured frequency and amplitude of the grouting hole in an impact-echo test would be considerably influenced by the hole depth.


Introduction
Grouted lap-splice connections are typical in precast concrete structures, which result in easy construction compared with other connecting systems (e.g., grouted sleeves). Generally, during the construction process, the connecting steel bars are inserted into the precast concrete components that need to be connected initially. ereafter, cement-based material (i.e., grout) is injected from the prepared hole (see "Grouting inlet" in Figure 1), and the connections are formed after the curing stage. Figure 1 shows the details of the grouted lap-splice connection.
Grouting quality directly affects the structural performance of the connected elements because it is the key part that connects two adjacent components together. If the grout contains large voids, the existing water and air inside the hole would cause corrosion of the connecting steel bars. Furthermore, it would influence the structural performance of the precast concrete components. In addition, defects in the grout will lead to stress concentration in the lapped bars, which results in a stress state that is inconsistent with the design and may affect the overall structural safety. erefore, it is necessary to develop a method to detect defects in grouted lap-splice connections.
Nondestructive testing (NDT) techniques have been developed to evaluate the material properties and soundness of concrete elements since 1940s [1]. e impact-echo (IE) method is one of the NDT techniques for concrete flaw detection. e IE method is based on the propagation and reflection of stress waves induced by a short-duration mechanical impact [1][2][3][4][5][6].
e IE response is related to the material property, the geometry of the test specimen, boundary conditions, flaw size, and flaw location within the specimen [1][2][3][4][5][6]. Details of the working principle of the IE testing method are introduced in the following section.
Although research has been conducted in this field, there is a lack of information on the application of the IE method in defect detection of grouted lap-splice connections.
erefore, the objective of this study is to investigate whether the existing IE method could be used to detect grouting defects in grouted lap-splice connections of shear walls. For this purpose, a precast shear wall containing grouting holes with a series of artificial defects was used to investigate the applicability of the IE method. e frequency response of the grouted holes with different defect levels were determined and compared in detail. In addition, finite element (FE) analyses were conducted to simulate the test and a parametric study was conducted to investigate the effect of hole depth on the frequency response.

eory of the IE Method.
e IE method is a NDT technique that incorporates the principle of low-frequency transient stress wave propagation. e principle of the IE technique is illustrated in Figure 2, in which T is the distance from the internal defect to the top plate surface and ∆t is the time when the P-wave first arrives at the signal receiving point. A small steel ball is used to impact the specimen surface, which generates stress waves that propagate through the test specimen. ree types of waves are generated, which are named as P-, S-and R-waves. e P-wave and S-wave are associated with particle motion parallel and perpendicular to the propagation direction, respectively. e R-wave (also known as Rayleigh wave) is a surface wave that travels along the solid surface. In the IE method, the measured displacement response is mainly influenced by the P-wave, owing to the close location of the sensor and impact point [1,2]. Reflection, refraction, and diffraction would occur when these waves encounter interfaces between materials with different acoustic impedances.
e sensor responds to the arrival of reflected waves, and the recorded time domain signals are transformed into the frequency domain using the fast Fourier transform (FFT) technique. e frequency with peak amplitude in the obtained spectrum represents the dominant frequency in the waveform.
is dominant frequency, also known as the thickness frequency, can be estimated as where K is a geometric correction factor that is determined by the geometrical shape of the structure [3]; n is a constant factor either equal to 2 or 4 depending on the acoustic impedances [3]; β is an empirical correction factor which has a value of 0.96 [1,2] or 0.95 [4,5]; and V p is the P-wave speed and is calculated through   Introduction. An IE test system mainly includes three components: a small steel ball, a transducer, and a waveform analyzer. e frequency content of the stress pulse is governed by the impact contact time, and most of the energy in a stress pulse exists in a frequency of 0 to 1.5/t c (t c is the contact time). With a decrease of the contact time, the frequency range would be increased. With shorter contact times, smaller defects can be discerned and shallower depths can be measured. However, the amplitude of each component frequency would be lowered, which decreases the penetrating ability of the stress waves [2]. As the contact time is a linear function of the ball diameter and has a very weak dependence on the kinetic energy at impact [2], the selection of the impact source significantly influences the results of an IE test. In this study, the impactor was a 6 mm steel sphere, which has a contact time of approximately 25 μs with the concrete [2].

Application of the IE Test
System. Whether a void in a slab or wall can be detected mainly depends on its lateral dimensions (l in Figure 3), depth (d in Figure 3), and the contact time [1]. If l > (1/4)d, the presence of the flaw can be detected. If l > (1/3)d, the flaw depth can be measured. If l > 1.5 d, the flaw behaves as an infinite boundary [1]. If (1/3) < (l/d) < 1.5 (shaded region shown in Figure 3), the amplitude spectrum typically has two peaks: a higher frequency peak corresponding to the depth of the flaw, which can be obtained only if the impact contact time is sufficiently short, and a lower frequency peak that would be lower than the thickness frequency of the unflawed plate. Figure 4 shows a typical precast shear wall panel as an example to clarify the above rules. In engineering practice, the grouting hole could be at different locations as shown in the figure. Given a grouting hole with the diameter of 35 mm, the hole depth in Figures 4(a) and 4(b) can be measured because the values of l/d are 0.7 and 0.467, respectively, which fall within the shaded region of Figure 3. If the grouting hole is located at the midthickness of the wall (Figure 4(c)), the depth could also be measured since l/d is 0.35 in this case. However, in the case of Figures 4(d) and 4(e), the l/d values are 0.28 and 0.233, which are lower than 1/3 and the depth cannot be measured. Furthermore, the flaw in the case of Figure 4(e) cannot be detected because the l/d value is 0.233, which is smaller than 0.25.

Test Specimen and
Instruments. An experimental investigation was conducted to study whether the IE method could be used to detect the voids in grouted lap-splice connections in a precast shear wall. e dimensions of the specimen are shown in Table 1. Details of the specimen can be found in Figure 5. ere are 18 typical grouting holes in the specimen with 9 holes (i.e., hole A through I) at each test surface (i.e., test surfaces 1 and 2). Here, the conditions of holes A, B, C, G, H, and I at test surface 1 were the same as test surface 2. Hole A consisted of solid grout with steel bars. Hole B consisted of only solid grout. Hole C had no grout but had steel bars. Holes G, H, and I consisted of defective grout and steel bars. Plastic foam was used to form defects of 10%, 20%, and 30% of the hole volume. Holes D, E, and F at test surface 1 had defects of 10%, 20%, and 30%, respectively, whereas at test surface 2, they were all solid grout with steel bars. A test on solid concrete of the shear wall was also conducted for reference. erefore, a total of 10 test conditions were considered in this study. e IE test system used in this study was SPC-MATS developed by Sichuan Central Inspection Technology Inc. as shown in Figure 6. A sampling frequency of 500 kHz and length of 4,096 points were used to capture the waveforms, for a record length of 8.192 ms. e frequency resolution in the computed amplitude spectra was 0.122 kHz. A steel ball with a diameter of 6 mm was used for impact. A piezoelectric accelerometer produced by Fuji Ceramics Corporation with a working frequency up to 15 kHz was adopted, and the electronic signals amplified were transferred into the digital signals through data acquisition and processing module. e test lines were arranged on both sides, directly above the holes. As shown in Figure 7, the measuring points were arranged at intervals of 40 mm along the test line direction. e distance between the impact point and the receiver location was between 20 and 30 mm. ere are 20 testing points along each hole; Points 1-15 are located directly over the hole, and Points 16-20 are located in the solid part to obtain the solid thickness frequency.

Test Results. Figures 8(a)-8(j)
illustrate examples of the waveforms and amplitude spectra of the 10 test conditions. Each figure represents one impact, and it should be mentioned that the results obtained from different impacts at one point remained almost the same. In Figure 8(j), it can be seen that the measured peak frequency of the solid wall (9.644 kHz) agrees well with the calculated value (9.50 kHz) using equation (1) while the V p is 3960 m/s. e test results obtained from Hole A were the same as the solid concrete l Depth can be measured Depth cannot be measured condition, including the time-history response and the frequency response, indicating that the hole A behaved as a solid infill material. For holes with grout defects, the predominant frequency did not change until a considerable defect occurred (i.e., hole F, H, and I). e empty hole C presented the lowest predominant frequency due to the hole (Figure 8(c)). is was because the propagation path of the stress wave was the longest due to no grout in the hole, which produced a longer travel time. Also, it should be mentioned that Figure 8(c) shows only one peak in the amplitude spectrum although two holes existed, which is because the frequency corresponding to the depth of the nearest hole was beyond the measuring range of the instrument. In addition, for some holes (e.g., Figure 8(e)), the obtained frequency might not indicate a defect although it was theoretically detectable, owing to the limited defect size compared with the size of the whole specimen. Taken together, the predominant frequency decreased with the increase of the void in the grout, indicating that the IE method offers the potential to detect defects in grouted lap-splice connections in precast shear walls.
In order to facilitate an in-depth understanding on the performance of IE method, the peak frequency obtained at each measuring point along each hole are shown in Figure 9, in which "− 1" and "− 2" represent results obtained at test surfaces 1 and 2, respectively. For hole A, the peak frequency values at the measuring points 1-15 were almost equal to the values at measuring points 16-20 (i.e., around 9.644 kHz), although slight scatter existed at some measuring points on surface 1 and they were also equal to the frequency values of hole B.
ese results indicate that the fully grouted holes behaved as solid concrete and the bars did not affect the obtained frequency value. However, for empty hole C, the peak frequencies at measuring points 1-15 were obviously lower than those at measuring points 16-20. Due to no grouting inside hole C, the propagation path of the stress wave increased, which decreased at the obtained peak frequency.
For hole G (10% void), the peak frequencies at measuring points 1-15 were slightly inferior to 9.644 kHz. However, the peak frequencies for holes H (20% void) and I (30% void) were as low as 9.277 kHz, which indicates that with the increase of the defect, the peak frequency would be reduced. Similar observations can be seen for holes D, E, and F. Here, it should be noted that the depth of the defect could not be discerned because the shifted frequency values were almost equal for the two test surfaces. However, the test results indicated that the IE method offered a good potential to indicate the relative level of the defect in the grouted lapsplice connections.

Finite Element Analysis
Two dimensional (2D) finite element (FE) analyses [19][20][21] based on the general purpose FE software ABAQUS were  Note. h g is the constructional measure height (see Figure 1), which is equal to the diameter of the grouting outlet hole in this study.
used to simulate the IE test. After a working model was developed, a parametric study was conducted to investigate the effect of the hole depth on the peak frequency in the amplitude spectrum. Details of the FE model, the comparison between test and analytical results, and the parametric study results are presented in this section.

General Description.
e view of the whole model is shown in Figure 10. Plane strain elements were used for modeling the concrete and the grout. In the model, perfect bond was assumed between concrete and grout (i.e., the "TIE" command in ABAQUS was used). In order to simulate the different defect levels, the grout shape was defined according the sectional view in Figure 5(d). e impact point is set above the defect, and the recording point is set 3 cm away from the    Table 2. In order to ensure the accuracy of the FE analysis, the element size was set as 5 mm [2]. e explicit dynamic analysis method was used. e impact duration was 25 μs. A load value of 5 N was adopted in the analysis. e time-history of the impact force is assumed to be a half sine curve. e record length is 6 ms with the sampling frequency of 500 kHz. e translation degrees of freedom at the left and right edges of the models in Figure 10 were fixed. Figure 11 shows the propagation of stress waves in the model under impact loading. After the analysis was completed, the peak frequency values from the test and analytical results were compared as shown in Table 3. It can be seen from the table that the analytical results agreed well with the test results. e experimental wave speed in solid part of the wall, which was calculated to be 4020 m/s by equation (1) error between the experimental and analytical frequencies was 2.83%. In the FE analysis, the ratio of frequency between hole C (empty) and hole A (solid) was 0.947, which is similar to the experimental value of 0.924. e higher value for the experimental peak frequency for the solid wall means that the actual concrete had a slightly higher wave speed than used in the FE model. ese observations indicated that the FE model could be used for modelling the frequency response of grouted lap-splice connections in the walls.

Parametric Analysis.
Based on the validated model, a parametric study was conducted to investigate the effect of the depth of an empty hole on the peak frequency response in the amplitude spectrum. A total of five depths were considered: 50, 75, 100, 125, and 150 mm. e hole diameter was 35 mm, and the thickness of the wall was 200 mm. e element size, boundary conditions, and material properties remained the same as the model mentioned above. e amplitude spectrums for the different hole depths are shown in Figure 12. e peak frequency (f 1 ), the frequency based on the reflection of hole (f 2 ), the corresponding amplitude, and the back-calculated depth by equation (1) are summarized in Table 4. Here, the wave path regarding f 1 and f 2 can be seen in Figure 12(f). It could be seen from the table that the back-calculated depth based on f 2 agreed well with the numerical model. However, the corresponding amplitude of f 2 was lower (See Figures 12(a)-12(e)) and it was hard to be obtained in engineering practice. In addition, the mode frequencies of the specimen can be seen in the frequency    spectrum. For example, in Figure 12(d), the peak value of 2.5 kHz was the ninth mode frequency of the specimen. e relationships between f 1 , the corresponding amplitudes and hole depth are summarized in Figure 13. It can be seen that as the hole depth increased, the peak frequency value initially decreased with depth until the hole was at the midthickness of the wall. en, the frequency increased with increasing depth. e response was symmetrical. However, the amplitude of the frequency peak decreased with increasing depth of the hole. is observation might help to further clarify the location of the void since the signal of f 1 was easy to obtain in engineering practice, which is essential in precast shear wall flaw detecting. However, it should be mentioned that the calibration was probably required to facilitate an accurate amplitude in test.

Conclusions
is study was mainly focused on investigating the feasibility of the impact-echo (IE) method in detecting defects in grouted lap-splice connections in precast walls. To this end, experimental investigations were conducted rstly. Eighteen grouting holes with di erent levels of arti cial defects were prepared in a shear wall specimen, and the IE method was used to obtain the frequency response data. In addition, nite element (FE) analyses were conducted to simulate the IE test, and a parametric study was conducted to investigate the e ect of the depth of an empty hole on the peak frequency value and its amplitude. e results obtained in this study allowed the following conclusions to be drawn: (1) e presence of a defect in the grout is indicated by a shift in the solid plate thickness frequency to a lower value. Based on the test results, the IE method demonstrated a good potential to detect grouting defects and could even provide a qualitative indication the defect level in grouted lap-splice connections. (2) e proposed FE model was con rmed to be suitable to simulate the dynamic response of the wall with grouting holes based on the comparison between test and analysis results. (3) Based on the parametric study, as the hole depth increased, the peak frequency decreased to a minimum value with the hole at the middle of the wall and then increased in a symmetric manner. e amplitude decreased continuously with increasing depth. (4) e parametric study also indicated that frequency shift to a lower value cannot be used as an indicator of the depth of the defect in engineering practice.
is study adopted a steel ball with a diameter of 6 mm used as impactor and an acceleration sensor, whereas those facilities are hard to catch high frequency response. In further study, di erent types of the facilities (e.g., tiny steel ball with conical displacement transducer) suitable for detecting the shallow defects discussed in this study will be adopted to validate the application of the IE method in grouting defect detection. In addition, in order to estimate the depth of the defect from the amplitude spectrum, a calibration on the obtained signal amplitude might be required in the future.
Data Availability e data used to support the ndings of this study are included within the article.

Conflicts of Interest
e authors declare that there are no con icts of interest regarding the publication of this paper.