**2. Implementation of AE approach for low-temperature cracking assessment of asphalt binders and asphalt concrete materials**

Low temperature cracking, a.k.a. thermal cracking, is a very common type of damage occurring in asphalt pavements located either in regions with cold climates or in milder climate regions with large daily temperature fluctuations. In asphalt pavements built in cold climates with severe winters, thermal cracking usually happens as a result of fast cooling rates (single-event thermal cracking). On the other hand, in asphalt roads located in regions with milder climate, thermal cracks develop at a slower rate, and it usually takes several cooling cycles for cracks to initiate and propagate through the pavement thickness (thermal fatigue cracking) [11]. When the temperature drops, surface of the pavement has the lowest temperature,

## *Application of Acoustic Emissions Technique in Assessment of Cracking Performance of Asphalt… DOI: http://dx.doi.org/10.5772/intechopen.101541*

and the temperature changes are highest there. Thermal tensile stresses develop in the restrained pavement layer due to the change in pavement temperature. The thermally-induced stresses are greatest in the longitudinal direction of the road which will lead to formation of transversely-oriented surface-initiated thermal cracks of various lengths and widths along the road.

Numerous research studies have demonstrated that the low temperature characteristics of asphalt pavements are closely related to that of the asphalt binder used in pavement construction. The AE method is implemented to evaluate the thermal cracking in asphalt binders. The AE binder sample consists of a 6 mm thick layer of asphalt binder bonded to an aluminum plate. To conduct the test, prepared specimens are placed inside the freezer and exposed to decreasing temperatures, ranging from 20°C to −40°C, or even to −50°C, if necessary for some polymer modified binders. To continuously monitor and record the sample temperature, a K-type thermocouple is placed on the specimens' surface. Due to the relatively small size of the AE sample, there is a thermal lag at the beginning of the test, which becomes negligible at temperature lower than −10°C. Differential thermal contraction between aluminum and asphalt binder induces progressively higher thermal stresses in the binder leading to formation of thermal cracks in the material. Thermal cracks formation in the sample is accompanied by a release of elastic energy in the form of transient waves which could be picked up using the AE piezoelectric sensor(s) mounted on aluminum plate. The critical cracking temperature, a.k.a. the embrittlement temperature, of the asphalt binders tested are determined by processing and analyzing the emitted elastic waves captured during the tests using the AE technique. **Figure 1a** schematically illustrates an AE testing sample of asphalt binder with an aluminum substrate [12–21].

To conduct the AE test for asphalt concrete materials, a semicircular-shaped asphalt concrete sample with a 50°mm thickness and a 150°mm diameter is used as the testing specimen, see **Figure 1b**. The testing sample for asphalt concrete can be fabricated from either field cores or from gyratory compacted samples. To conduct the AE test, similar to the binder test, the prepared AE sample is subjected to decreasing temperatures ranging from 20°C to -40°C and the acoustic activities and temperature of asphalt material test sample is continuously monitored and recorded using piezoelectric AE sensors and a K-type thermocouple, respectively. The source of acoustic emission activities in asphalt concrete materials is formation of thermally-induced microdamages within the asphalt mastic. As a heterogeneous viscoelastic material, the thermally induced stresses develop in asphalt concrete due to the thermal contraction mismatch between aggregates and surrounding asphalt mastic [12–26].

The acoustic emission testing set up used for assessing asphalt materials consists of several wideband piezoelectric AE sensors along with pre-amplifiers and data acquisition system with processing and analysis software. The Digital Wave-Model B1025 wideband AE sensors with nominal frequency range of 20 kHz to 1.5 MHz used in this study in order to record and continuously monitor AE activities of the material while conducting the experiment. To reduce extraneous noise, the AE signals picked up by AE sensors are first pre-amplified 20 dB using broad-band preamplifiers. Then AE signals are amplified again 21 dB for a total of 41 dB. At the end signals are filtered using a 20 kHz high-pass double-pole filter through using the signal conditioning unit. A 16-bit analog to digital converter (ICS 645B-8) with 2 MHz sampling frequency and a length of 2048 points per channel per acquisition trigger are used to digitize the signals and outputs are stored for the post-processing.

In general there are two methods normally used to analyze AE signals: (1) the "classic" or "parameter-based" method; (2) the "quantitative" or "signal-based" approach. In the first approach, the AE signals are not recorded, instead only some

#### **Figure 1.**

*AE testing samples used for (a) asphalt binders (b) asphalt concrete materials.*

AE parameters are recorded and analyzed. Whereas in the signal-based approach, the actual AE signals are recorded and used to analyze the materials microstructure. The failure and microdamage occurring in the material could generate significant number of AE signals within a very short time generating big amount of AE data. In the parameter-based method only some rudimentary analysis can be performed on AE data however it is faster than the quantitative method. On the other hand, while the signal-based approach is slower, it is capable of more sophisticated analysis of performance of the material.

For evaluation of thermal cracking in asphalt materials, both parameter-based and signal-based techniques were implemented on recorded AE signals and associated test temperature. AE event is an individual waveform with the threshold of 0.1 V and the energy level equal to or greater than 4 V2 μs. The emitted energy associated with each event is one of the important characteristics of an AE signal and can be calculated using Eq. (1), where EAE is the AE energy of an event (V2 μsec) with duration of time t (μsec) and recorded voltage of V(t) [1].

$$\mathbf{E}\_{AE} = \underset{\mathbf{0}}{\overset{\text{t}}{\mathbf{V}}} \mathbf{V}^2 \left(\mathbf{t}\right) dt \tag{1}$$

**Figure 2** shows a typical plot of AE events counts versus temperature for typical asphalt binder and asphalt concrete AE tests which consists of four distinct

## *Application of Acoustic Emissions Technique in Assessment of Cracking Performance of Asphalt… DOI: http://dx.doi.org/10.5772/intechopen.101541*

regions, namely: (1) pre-cracking, (2) transition, (3) stable cracking, and (4) fully cracked regions. In the "pre-cracking region", thermally-induced stresses in the sample are building up and they are still below the strength of the material. As a result no damage and consequently no AE events are observed within this region. In the "transition" region, as soon as the thermal stresses reach the strength of the material, microdamages form in the material which manifests itself as a cluster of high amplitude AE events. The temperature corresponding to the AE event with the first peak energy within the transition region has been termed the "embrittlement temperature," as shown in **Figure 3**. The embrittlement temperature is the onset of damage in asphalt material. Results has demonstrated that the embrittlement temperature is a fundamental material state which is independent of material constraint, sample size (as long as a statistically representative volume or larger is used), and sample shape [15]. In the "transition region", material behavior gradually changes from a quasi-brittle to a brittle state where resistance to fracture is generally very low, allowing microdamages to propagate readily.

The third region is the "stable cracking region" which normally initiates at a very low temperatures when the material is brittle. Significant amount of AE activities are observed during this region. The last region, is the "fully cracked region" where the rate of AE activities of the sample begins to reduce until it reaches almost zero at the end of this region. The AE activities originate from formation of new microdamage inside the sample. Thus reduction in the rate of AE activity can be linked to the presence of plenty of microdamage in the sample. This region is usually observed when the sample is cooled down to very cold temperatures allowing all microdamage to develop within the sample [15].

**Figure 4** illustrates the typical envelope locus of AE event energies of asphalt samples and demonstrates the intensity of the released energies of AE events. In the pre-cracking region, the envelope locus is zero and suddenly at the beginning of the transition region it jumps to its maximum magnitude. The magnitude of AE event energies gradually tapers off in stable cracking region until it reaches almost zero in the fully cracked region.

The histogram presented in **Figure 5** shows the graphical representation of the distribution of AE events energies for asphalt materials. Results suggest that only a small portion of AE events are high energy events while the rest of the events are in fact low energy. Generally, the energy content of an event is proportional with the size of the microdamage causing that event. The high energy events result from the formation of large microcracks while the low energy AE events could be linked to formation of hairline microcracks in the material.

**Figure 2.** *Typical AE event counts vs. temperature plot regions.*

**Figure 3.** *Typical plot of event count and AE energy vs. temperature [15].*

**Figure 4.** *Typical envelope locus of AE events energy during thermal cooling [15].*

The AE test results for 24 different types of asphalt materials (eight different binders, each at three aging levels) including: AAA-1 (PG 58-28), AAB-1 (PG 58-22), AAC-1 (PG 58-16), AAD-1 (PG 58-28), AAF-1 (PG 64-10), AAG-1 (PG 58-10), AAK-1 (PG 64-22), AAM-1 (PG 64-16) are presented in **Figure 6**. In this experiment each binder was tested at three aging levels: (1) unaged (TANK) (2) short-term aged (RTFO), and (3) long-term aged (PAV). The ASTM D2872-04 (ASTM 2004) and ASTM D6521-05 (ASTM 2008) were used to perform the oxidative aging process of RTFO and PAV binders, respectively. It should be mentioned that in PG XX-YY used for expressing the Performance Grade of asphalt materials, XX corresponds to the expected average high temperature of asphalt pavement over a 7 days, and YY is the lowest expected temperature of the pavement.

Results show that the AE embrittlement temperatures correlated well with the bending beam rheometer (BBR-based) critical cracking temperatures with R2 = 0.85. Results suggest that AE-based embrittlement temperatures are lower than the corresponding BBR-based critical cracking temperatures. This could be *Application of Acoustic Emissions Technique in Assessment of Cracking Performance of Asphalt… DOI: http://dx.doi.org/10.5772/intechopen.101541*

**Figure 5.** *Typical histogram of AE events energies for asphalt binder material [15].*

#### **Figure 6.**

*Correlation between AE embrittlement temperature and BBR-based cracking temperature illustrating the conservative nature of the BBR-based cracking temperatures [21].*

attributed to the fact that the AE-based embrittlement temperatures are directly related to the cracking performance of the material while the BBR- base critical temperatures are based upon the binder's rheological material properties and include an inherent factor of safety to avoid low-temperature pavement cracking. In addition, numerous studies have demonstrated that AE approach is sensitive to aging level of the material and could successfully evaluate asphalt materials at different oxidative aging levels. Finding of different studies show that the embrittlement temperature of asphalt materials is sensitive to aging levels, where TEMB-TANK < TEMB-FTFO < TEMB-PAV.
