Initially, twenty-four rock samples were prepared using large size blocks of banded amphibolite, which were collected from varied rock slopes along Chengdu-wenchuan highway in Sichuan province in china for laboratory investigation. These rocks were selected due to their high rock-fabric complexity and their high significance impact in rock engineering design. Macroscopically, the rocks exhibit a distinct planar syn-metamorphic fabric that characterized with near perfect tectonic metamorphic layering and mineral alignment/elongation (schistosity) and penetrative stretching lineation. The foliation bands varying in thickness from 1 to 3 mm and stretched to a length of 5 mm (Figure 1(a)).

For the laboratory test, the collected rock blocks were trimmed with their side’s perpendicular to each other to facilitated coring in at different angles. A total of eighty five cylindrical specimens of length to diameter ration equal to 2 were cored at different orientation of anisotropy planes β (0˚, 30˚, 60˚ and 90˚) (Figure 1(b)). Rock specimens were cored using conventional laboratory drilling machine and the ends trimmed as required as per the [18] standards. After coring, the rock specimens were further smoothened using a lathe to avoid end effects. The specimens were dried at 105˚C for 24 h to remove the moisture. The density is determined for core specimens following the standard test procedures outlined in [18] (Table 1).

Obtaining a reliable relationship is contingent on an effective petrographic study of the rockcore samples. Thus, to evaluate the petrography and petro-fabric of the studied rocks, thin sections across and along the schistosity were prepared and observed under a high power polarized microscope. The mineral assemblage and texture

characteristics of the typical rocks may vary within the meter scale. However, slight compositional heterogeneity may found down to the centimeter scale. Figure 2 shows the microscopic textural variations of visually selected rock subtype samples of amphibolite rocks, which is considered as amphibolite schist based on its mineral assemblage of low-grade metamorphism. Although it can be classified as amphibolite gneiss texturally based on the alternative felsic and mafic minerals, which is preserved in the gneissosity texture.

The predominant mineralogical composition is almost hornblende, actinolite, quartz and feldspar with minor amounts of epidote, biotite, chlorite and calcite. Microscopic descriptions of thin sections show that the rock samples contain varying proportions of nematoblastic hornblende set in a fine-to medium-grained quartz-pla- gioclase mosaic. Furthermore, the main microstructure is characterized by well-developed layered lattice structure that defined by strong crystallographic preferred orientation of hornblende (Figure 2(a) and Figure 2(b)) and elongated quartz grains in quartz rich zone (Figure 3(c) and Figure 3(d)). Aligned xenoblastic hornblende crystals form the bulk of the rock that average is 0.15 mm, although large crystals reach 0.3 mm, and they usually contain numerous small inclusions of quartz (Figure 2(a) and Figure 2(c)).

Apparently, the orientation of the mineral inclusions is generally controlled by the cleavage of the host hornblende. In some thin sections, numerous porphyroblasts of quarts and plagioclase considerably deflect the schistosity plane (Figure 2(b)). Prismatic (Hornblende and actinolite) minerals are occurring in alternate order with respect to quartz felspathic bands (Figure 2(a)). The overall symmetry of the hornblende fabric is orthorhombic with the foliation as the symmetry plane. The alignment of the rod-shaped hornblende minerals defines the lineation within the foliation plane (Figure 2(d)). Some parts of fine-grained amphibolite composed almost entired of small granular epidote crystals and the hornblende has altered to stellate aggregates of fibrous actinolite and chlorite (Figure 2(f)). Granular epidote crystals are concentrated in parallel bands upto 1 mm. thick or occurred in sub circular or elliptical shape (Figure 2(e) and Figure 2(f)). The presence of chlorite and epidote along with the aforementioned mineral assemblage suggests that the rock belongs to low grade regional metamorphic facies. The accessory minerals include calcite, apatite, sphene and iron oxide.

Ultrasonic test is used as a method to determine the velocity of propagation of P and S waves in laboratory cylindrical core samples and then to determine the elastic symmetry of anisotropic rocks. In this technique, the frequency of wave should be high and rock specimens should have infinite extent compared to the wavelength of the pulses. Twenty-four cylindrical rock samples were prepared by cutting and lapping the ends and the length is measured for each. An ultrasonic digital indicator consist a pulse generator unit, transmitter and receiver transducers are used for sonic pulse velocity measurement. The transmitter and the receiver are positioned at the ends of each specimen and the travel time of the pulse wave is measured. The velocity is calculated by dividing the length of rock sample by wave travel time.

Both velocities of the compressional P-wave and shear S-wave were measured for each core samples with reference to the foliation angle and the average of 3 to 5 sample results was determined (Table 1). Figure 3 shows the used instrument for the P- and S-velocity measurements based on ASTM [19] standard. The velocity of ultrasonic pulses travelling in a solid material depends on the density and elastic properties of that material.

From the theory of elasticity, compressional (or longitudinal) P-wave velocity (V_p) is related to the elastic modulus (E), and the density (ρ) of the material as in Equation (1). If ρ in g/cm³ and V_p in km/s, then E in GPa (10⁹ N/m²). The elastic modulus estimated by this method is the sometime termed as seismic modulus (also called dynamic modulus, but should not be mistaken as the modulus under dynamic compression). It is different from the modules obtained by the uniaxial compression tests. The value of the seismic modulus is generally slightly higher than the modulus determined from static compression tests [20]. Similarly, seismic shear modulus G may be determined from shear S-wave velocity Vs (Equation (2)). Gsis in GPa, when density ρ is in g/cm³, and S-wave velocity Vs is in km/s. Seismic Poisson’s ration ν can be determined from Equation (3). Alternatively, seismic Young’s modulus Es can be determined from shear modulus (G_s) and Poisson’s ratio (νs) as in Equation (4).

In order to examine the influence of velocity anisotropy of banded amphibolite the results obtained from the measurement of wave velocities in different directions to the foliation planes are summarized in Table 1. The P-wave velocities range from 3952.6 to 5441 m/s with mean 4676.9 m/s and for S-wave velocity varied between 2340.8 and 3178.9 m/s with mean value 2842.3 m/s. The average correlation of P- and S-wave follows the expression V_p = 1.6 Vs. The wave velocities with orientation of foliation planes (0˚, 30˚, 60˚, and 90˚) relationship are characteristic for the investigated rock specimens (Figure 4). Because elastic wave propagation is very sensitive to weak discontinouty, velocities propagate parallel to the oriented fractures are much higher than those in perpendicular propagation [21]. It is obvious that the planes of foliation have a major effect on the wave velocity, where the V_p and V_s taken parallel to the foliation plane show highest values than those obtained in the other directions (β = 30˚, 60˚ and 90˚). However, it was detected in some core samples varied values of V_p and V_s with in the same orientation plane due to the numerous porphyroblasts of quarts and plagioclase that considerably deflect the schistosity plane as shown in Figure 2. At the scale of microscope, it may be observed that these rocks are characterised by a marked orientation of grains planes systematically oriented parallel to foliation.

The anisotropy degree (defined as the ratio of the maximum to the minimum wave velocity) is equal to 1.38 and 1.36 for the longitudinal (P) and transverse (S) wave velocity respectively. The behavior of the sebanded amphibolitesis similar with observations made by Saroglou et al. [22] who tested different gneissic rocks (from Egnatia highway, in northern Greece) parallel and perpendicular to foliation. They determined a mean compression and shear velocity of V_p = 4182 m/s and V_s = 2668 m/s respectively, parallel to foliation, while perpendicular to it the velocity was V_p = 2608 m/s and V_s = 1578 m/s. This was in accordance with the findings of the current research (Table 1).

Although values of elastic constants properties, by ultrasonic evaluation, tend to be higher than those preliminary predicted by static laboratory methods, however these values are nonetheless a useful in preliminary prediction of static properties [23]-[28].

However, in this study, a correlation analyses were carried out in order to study the influence of the orientation of foliation planes on the elastic constants that obtained from ultrasonic pulse test. Figures 5-7 shows the relationship results between the three measuring elastic constants and dip angle (β). There was significant correlation found between Young’s modulus (E) and shear modulus against dip angle with correlation coefficient R² = 0.61 - 0.82 and R² = 0.56 - 0.82, respectively (Figure 5 and Figure 6) except the Poisson’s ratio that shows poor correlation coefficient (R² = 0.01 - 0.3) to the orientation of the foliation planes (Figure 7). The result indicates that the maximum values of the E and G are for a β of 0˚ and minimum values for β of 90˚ therefore no clear

dependence on the mineralogical and textural varieties of studied rocks was experienced. It is summarized that the observed intrinsic anisotropy and the close relations of the directional dependent seismic anisotropy to the foliation planes is mainly a result of crystallographic preferred orientation of major minerals (e.g. hornblende and elongated quarts grains).