Influence of Mineralogical and Petrographic Properties on the Mechanical Behavior of Granitic and Mafic Rocks

Abstract

This study investigates the impact of mineralogical and petrographic characteristics on the mechanical behavior of granitic and mafic rocks from the Shuangjiangkou (Sichuan Province) and Damiao complexes (Hebei Province) in China. The research methodology combined petrographic investigation, comprising optical microscopy and Scanning Electron Microscopy–Energy-Dispersive X-ray Spectroscopy (SEM-EDS) methods, with methodical geotechnical characterization to establish quantitative relationships between mineralogical composition and engineering properties. The petrographic studies revealed three lithologic groups: fine-to-medium-grained Shuangjiangkou granite (45%–60% feldspar, 27%–35% quartz, 10%–15% mica), plagioclase-rich anorthosite (more than 90% of plagioclase), and intermediate mangerite (40%–50% of plagioclase, 25%–35% of perthite). The uniaxial compressive strength tests showed great variations: granite (127.53 ± 15.07 MPa), anorthosite (167.81 ± 23.45 MPa), and mangerite (205.12 ± 23.87 MPa). Physical properties demonstrated inverse correlations between mechanical strength and both water absorption (granite: 0.25%–0.42%; anorthosite: 0.07%–0.44%; mangerite: 0.10%–0.25%) and apparent porosity (granite: 0.75%–0.92%; anorthosite: 0.20%–1.20%; mangerite: 0.29%–0.69%), with positive correlations to specific gravity (granite: 1.88–3.03; anorthosite: 2.67–2.90; mangerite: 2.43–2.99). Critical petrographic features controlling mechanical behavior include the following: (1) mica content in granite creating anisotropic properties, (2) extensive feldspar alteration through sericitization increasing microporosity and reducing intergranular cohesion, (3) plagioclase micro-fracturing and alteration to clinozoisite–sericite assemblages in anorthosite creating weakness networks, and (4) mangerite’s superior composition of >95% hard minerals with minimal sheet mineral content and limited alteration. Failure mode analysis indicated distinct patterns: granite experiencing shear-dominated failure (30–45° diagonal planes), anorthosite demonstrated tensile fracturing with vertical splitting, and mangerite showed catastrophic brittle failure with extensive fracture networks. These findings provide quantitative frameworks that relate petrographic features to engineering behavior, offering valuable insights for rock mass assessment and engineering design in similar crystalline rock terrains.

1. Introduction

The mechanical behavior of intact rock is a crucial aspect in various geotechnical and geoengineering practices, particularly in deep underground space exploration and development [1,2]. Tunnels, mines, hydropower plants, and nuclear waste repositories are projects where a clear understanding of rock strengths, deformation, and failure is critically important to ensuring safety and stability [3,4]. Accurate and rapid estimation of rock mechanical properties, such as uniaxial compressive strength (UCS), is crucial for effective design and reduces project costs. However, obtaining intact rock cores to perform normal laboratory tests can be challenging, particularly in complex geological settings such as deep mines, where the rock masses are fractured [2,5,6,7,8,9]. This highlights the need for approaches that describe rock mechanical properties in terms of readily determined intrinsic rock properties.

The mechanical behavior of rock materials is always complex and depends on many parameters grouped into intrinsic geological properties and extrinsic environmental or stress conditions [10,11,12]. The intrinsic properties include mineralogical composition, grain size and shape, texture (grain contacts, arrangement, presence of microcracks), porosity, and the extent and nature of alteration or weathering [3,11,13,14,15,16,17,18,19,20]. These properties define the natural rock’s strength and its response to stresses imposed. Many studies have been conducted on the correlation between geological properties and the mechanical behavior of different types of rocks, particularly granites and sandstones, which are commonly found in engineering projects [3,4,10,14,17,21,22,23,24]. Parameters, including modal mineralogy, grain size, and textural features, are correlated with mechanical strength parameters such as UCS and tensile strength in studies conducted through petrographic microscopy [10,16,25]. The proportion of minerals such as quartz and feldspar, their relative ratios, and grain size are major factors that affect strength. In general, strength can be enhanced by a greater quartz content or quartz/feldspar proportion. In contrast, the quantity of such minerals as plagioclase, biotite, or muscovite can decrease strength [17]. Grain size is also critical, with smaller grain sizes being generally linked with greater strength in granitic rocks [10]. The textural aspects of the arrangement and interlocking of grains and the type and length of grain contacts are also known to affect rock mechanical properties [26,27].

In addition to mineralogy and texture that are observable by standard petrography, the physical properties of rocks, especially porosity and capacity to absorb water, are very important in determining the durability and vulnerability of rocks to degradation [28]. Porosity is the total amount of space in the rock, and water absorption is the capacity of the rock to absorb water, filling the pores and microcracks. Pores and absorbed water can significantly reduce the effective load-bearing area, facilitating physical processes of weathering, such as freeze–thaw damage, and altering intergranular forces [29]. The rocks with low porosity and water absorption tend to be stronger and more durable. Another basic physical property that is associated with the bulk mineral composition and compactness of the rock is specific gravity (or density) [28]. Although a traditional petrographic analysis is informative in terms of the modal composition and texture of the rock, more in-depth information on the intrinsic properties of the rock demands sophisticated techniques of characterization [1]. Mineralogical tools, including Scanning Electron Microscopy (SEM) and Energy-Dispersive X-ray Spectroscopy (EDS), provide effective ways to study the fine mineralogical composition, crystalline structure, micro-textural features, grain boundary nature, existence and type of microcracks, and density of alteration minerals [30]. SEM provides very sharp images of the rock microstructure, revealing important details such as intergranular contacts, microcrack networks, and the shape of mineral grains and alteration. EDS can be used to determine the exact mineral phases, both minor and altered, which may be hard to distinguish by optical microscopy [31]. Such micro-scale characteristics and composition details are essential for the initiation and propagation of fractures, which ultimately dictate the macroscopic mechanical behavior and failure modes of the rock [32,33]. As an example, the distribution and abundance of weak mineral phases or pre-existing microcracks, observed through SEM and defined using EDS, may provide stress concentration sites and fracture initiation and propagation pathways during loading [34].

Although the correlations between basic petrographic parameters, standard physical properties, and mechanical strengths are studied relatively well in common rocks such as granite and sandstone, more comprehensive studies are needed that combine detailed microstructural and compositional characterization (using SEM and EDS) with petrophysical properties (water absorption, porosity, specific gravity) and fundamental mechanical behavior (UCS) over a wider range of rock classes, especially mafic rocks. Moreover, the analysis of rocks, particularly in geologically distinct places, can provide a significant amount of context about the role of formation history and local geological processes in creating differences in intrinsic properties and thus mechanical behavior. This study is based on two major types of rocks in important geological and engineering areas in China: granitic rocks in Shuangjiangkou, Maerkang County, Sichuan Province and mafic rocks, including anorthosite and mangerite, of the Damiao complex in Chengde County, Hebei Province. The Damiao complex is a well-constrained Paleoproterozoic massif-type anorthosite complex that is situated in the northern North China Craton [35,36]. The Damiao complex was emplaced between 1.75 Ga and 1.68 Ga, providing a distinctive geological environment in which complex magmatic and post-magmatic processes have led to large compositional and textural variations [37,38,39,40]. The mechanical behavior of such diverse igneous rocks requires a multi-faceted approach that incorporates detailed characterization on a variety of scales and property domains. This research aims to perform a comprehensive study of the effects of mineralogical and petrophysical properties on the mechanical behavior of granitic and mafic rocks, with a specific aim of characterizing the detailed mineralogical composition and micro-texture through optical microscopy, SEM, and EDS, determining key physical properties of water absorption, porosity, and specific gravity, measuring the uniaxial compressive strength and establishing relationships between the measured properties and mechanical strength of both rock types.

2. Geological Conditions of Study Area

2.1. Shuangjiangkou Granite

The Shuangjiangkou hydropower station is situated in Dadu River Basin, Sichuan Province in China in a tectonically active area on the boundary of the Eurasian and Pacific tectonic plates [41]. The geological terrain consists of high mountains and deep-cut valleys, resulting from crustal movement and the erosion of the surface by weathering. The major type of rock is Yanshanian porphyritic biotite-K-feldspar granite, which has typical hard and brittle rock properties [41,42,43], and other varieties are biotite K-feldspar granite and porphyritic K-feldspar granite (Figure 1). Petrographic analysis indicates a medium-fine to fine-medium grained structure with 50%–56% feldspar, 26.5%–35% quartz, and 10%–15% mica with small amounts of zircon and iron [44]. The quartz volume is high, and this is a major contributor to the brittle failure nature, whereas the mica has an intermittent directional distribution nature.

The in situ stress conditions are exceptionally high in the study area, with the measured principal stresses of 16–38 MPa (σ₁), 9–20 MPa (σ₂), and 3–10 MPa (σ₃) in underground cavern regions at burial depths of 2–6 km [43,44]. The rock mass structure is mainly blocky, and there are no regional faults in the area of the underground caverns. Nevertheless, secondary inactive faults form the primary structural planes, which include certain discontinuities, such as fault SPD9-f1 [45]. These structural characteristics, along with high geostress and weathering factors, significantly impact the behavior of the rock mass. Thus, construction activities can lead to excavation-induced failures, such as spalling and slip-type rockbursts.

Figure 1. Overview of the Shuangjiangkou site: (a) location and topography; (b) geological profile at EL.2279.90 m (revised from [46]).

Figure 1. Overview of the Shuangjiangkou site: (a) location and topography; (b) geological profile at EL.2279.90 m (revised from [46]).

2.2. Chengde Anorthosite

The Damiao anorthosite complex is in the northern part of the North China Craton [47,48,49] (Figure 2) in Chengde County, Hebei Province, China. It is a Proterozoic massif-type anorthosite complex of the 1.75 Ga–1.68 Ga suite of anorthosite–mangerite–alkali granitoid–rapakivi granite intrusions, and the Damiao complex is dated at around 1.74 Ga [39,48]. The complex formed due to an ancient lower crustal source and was intruded at mid-crustal depth at a post-collisional tectonic setting associated with a Paleoproterozoic orogen [36,49]. The major rock type is anorthosite (~85%), and minor lithologies are leuconorite, melanorite, mangerite, oxide–apatite gabbronorite, perthite noritic, and ferrodiorite [36,39]. Gabbroic and ferrodioritic dikes crosscut the complex, and it has a rhythmic anorthositic–pyroxenitic layering [50].

The anorthosite is represented by two different varieties: altered, white-colored, and weakly to non-altered, dark-colored anorthosite (Figure 3). The dark color is due to the high content of FeTi oxide inclusions in plagioclase crystals. The majority of anorthosite has experienced broad hydrothermal alteration (auto-metamorphism) to yield a high-friability, low-density, white-colored anorthosite that outcrops in an area of more than 60 km². In the white anorthosite, the dark-colored anorthosite is found as rare, irregular blocks (1–50 m²) with gradational contacts having transitional grayish zones in which gray-black fresh plagioclase is found at the core of altered white plagioclase. The alteration process implied decreases in CaO and MgO and increases in SiO₂, Al₂O₃, and Na₂O, which proved that white anorthosite was formed by weathering and hydrothermal alteration of the initial dark one. No important fault systems or stress conditions are mentioned in this locality; the complex is mainly influenced by regional weathering and hydrothermal alteration processes.

2.3. Chengde Mangerite

The mangerite bodies were emplaced as small volumes into the anorthosite in the northwestern component of the Damiao complex in North China Craton (Figure 2). The mangerite contains plagioclase, K-feldspar, quartz, and minor amounts of clinopyroxene, FeTi oxides, apatite, and hornblende [35]. The mangerite occurrences are not influenced by any major fault systems or stress regimes, and weathering processes are the key geological processes to influence the intermediate composition rocks.

3. Materials and Methods

3.1. Specimen Preparation

The samples of homogeneous granite, anorthosite, and mangerite, with very limited natural fracturing, were chosen in Maerkang County, Sichuan Province and Chengde County, Hebei Province in China, respectively. The selection criteria focused on intact rock masses with uniform textures and no visible discontinuities to provide representative results of mechanical testing. The specimens were obtained in standard cylindrical form using the specimen preparation guidelines proposed by the International Society for Rock Mechanics (ISRM) [53] by carrying out coring, cutting, and accurate grinding techniques to obtain the specimens. All specimens were manufactured to a nominal diameter of 50 mm and to a height-to-diameter ratio of 2:1 in accordance with standard testing guidelines. A lot of care was taken to achieve geometrical accuracy during the preparation work, and the non-parallelism and non-perpendicularity of the end faces of the specimen have been kept within strict tolerance below 0.02 mm, as shown in Figure 4.

3.2. Engineering Properties

The granitic and mafic rocks were geotechnically analyzed using an in-depth analysis of physical and mechanical characteristics. The basic rock properties were characterized by determining physical properties such as the water absorption, apparent porosity, and bulk density of the rock according to the standard procedures. To determine how rock composition and structure affect mechanical behavior, uniaxial compression testing was employed to examine mechanical properties, specifically compressive strength. The experimental program was conducted using an Instron 8800 Hydraulic Servo Fatigue Testing Machine (Instron, Norwood, MA, USA), as indicated in Figure 5. The technical facilities of this apparatus include advanced capabilities, such as a high-stiffness servo-controlled loading frame, adaptive coordinated feedback control, and accurate measurement systems, which enable the recording of the process of fracture development and deformation [54]. The high structural stiffness is key to ensuring that the full three-dimensional stress–strain response of brittle rock specimens is measured, especially in post-peak behavior when specimens lose their load-bearing capacity and undergo stress drops. The solid loading frame limits the deformation of the apparatus during the test, thereby limiting the amount of energy applied externally and maintaining data integrity throughout the experiment. The uniaxial tests were conducted using a specific procedure, depending on the capabilities of the testing equipment. The diameter and height of each cylindrical specimen were measured and recorded before testing. A 12.5 cm heat-shrinkable tube was sealed to the specimen to preserve moisture content and field conditions until testing, and both axial and circumferential extensometers were used to measure deformation. The metrological parameters of the measurement system were as follows: load cell with a measurement range of 0–500 kN, measurement resolution of 0.01 kN, and measurement uncertainty of ±0.5% of full scale; axial extensometers with a measurement range of ±6.25 mm, measurement resolution of 0.001 mm, and measurement uncertainty of ±0.1% of reading; and circumferential extensometers with a measurement range of ±6.25 mm and measurement resolution of 0.001 mm; measurement linear variable differential transformers (LVDTs), compressometers, and electrical resistance strain gauges were employed to make accurate measurements, and the sensor ranges were regulated to −6.25 mm to 6.25 mm and initial readings to around −4 mm. The tests were conducted under controlled environmental conditions at room temperature (20 ± 2 °C) and relative humidity of 50 ± 5%. All specimens were tested under the same environmental conditions to ensure consistency in the testing conditions and minimize the impact of environmental factors on mechanical properties. The prepared sample was then inserted in the instrument. A small pre-load of 1 kN was applied under force control to ensure that firm seating and appropriate contact were made between the loading platens and the specimen. Once the pre-load was reached, the system was converted to displacement control, and a constant rate of axial displacement of 0.06 mm/min was imposed until the specimen failed. In the process, the axial load, axial deformation, and circumferential deformation were continuously monitored to obtain the compressive strength of the rock and complete stress–strain curve.

The physical properties (water absorption, apparent porosity, and specific gravity) were determined by the following standard equations [55] to provide a baseline material characteristic against which the mechanical performance parameters can be correlated.

$$\mathrm{Water} \mathrm{absorption} (\%)={\displaystyle \frac{(\mathrm{S}\mathrm{S}\mathrm{D} \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}-\mathrm{D}\mathrm{r}\mathrm{y} \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t})}{\mathrm{D}\mathrm{r}\mathrm{y} \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}}}\times 100$$

(1)

$$\mathrm{A}\mathrm{p}\mathrm{p}\mathrm{a}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{t} \mathrm{p}\mathrm{o}\mathrm{r}\mathrm{o}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{y} (\%)={\displaystyle \frac{(\mathrm{S}\mathrm{S}\mathrm{D} \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}-\mathrm{D}\mathrm{r}\mathrm{y} \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t})}{(\mathrm{S}\mathrm{S}\mathrm{D} \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}-\mathrm{S}\mathrm{u}\mathrm{s}\mathrm{p}\mathrm{e}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{d} \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t})}}\times 100$$

(2)

$$\mathrm{S}\mathrm{p}\mathrm{e}\mathrm{c}\mathrm{i}\mathrm{f}\mathrm{i}\mathrm{c} \mathrm{G}\mathrm{r}\mathrm{a}\mathrm{v}\mathrm{i}\mathrm{t}\mathrm{y} (\mathrm{k}\mathrm{g}/{\mathrm{m}}^3)={\displaystyle \frac{\mathrm{D}\mathrm{r}\mathrm{y} \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}}{(\mathrm{S}\mathrm{S}\mathrm{D} \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}-\mathrm{S}\mathrm{u}\mathrm{s}\mathrm{p}\mathrm{e}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{d} \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t})}}\times 1000$$

(3)

3.3. Petrographic Characterization

To determine the rock-forming minerals, grain textures, and alteration or deformation characteristics, the petrographic characterization was performed by detailed optical microscopy analysis (Figure 6A). The thin sections were polished at the Institute of Geology and Geophysics, Chinese Academy of Sciences (CAS), Beijing, under conventional methods of petrography. The sample rocks were first cut into slabs of a few millimeters in thickness with a diamond saw-fitted rock-cutting machine and then polished using a rotary grinding machine to obtain smooth surfaces. Slabs were prepared, mounted on Canada Balsam adhesive on glass slides, and ground progressively using finer abrasive grits until the standard thin section thickness of ca 0.03 mm was achieved. The mineral assemblages, grain size distributions, and fabric orientations were characterized under both plane-polarized and cross-polarized light using optical microscopy.

3.4. SEM-EDS Mineralogical Analysis

The mineralogical and elemental analysis was carried out with a TM4000plus Scanning Electron Microscopy unit (Hitachi, Tokyo, Japan) fitted with Bruker Quantax 75 Energy-Dispersive X-ray Spectroscopy (EDS) at the Institute of Geology and Geophysics, Chinese Academy of Sciences (IGGCAS), as seen in Figure 6B. Back-scattered electron (BSE) imaging, performed at an accelerating voltage of 20 kV, was used to demonstrate the textural characteristics and compositional differences between quartz and feldspar minerals, with each image taking approximately one minute to acquire. At 20 kV, semi-quantitative spot analyses were carried out to determine elemental composition and validate mineral identifications. Single spot analysis took about 15 s, and the area of the spectrum was more than 2.5 × 10⁵ counts for analytical accuracy.

4. Results

4.1. Mechanical Behavior of Rocks

The mechanical behavior of crystalline rocks is determined by complex interactions between physical properties (absorption of water, porosity, specific gravity) and petrographic properties (mineral composition, texture, alteration, microfractures). The main control is performed by mineral composition, by differences in mechanical properties of the constituent phases. Mica, especially biotite, shows strong negative correlations with the strength parameters because of the perfect basal cleavage and low stiffness [56]. Internal stress concentrations formed by textural features such as grain size and grain distributions localize deformation. The populations of microfractures are essential characteristics that determine the mechanical behavior where the pre-existing discontinuities serve as the nucleation sites of failures [57]. Mineral alteration is a fundamental process that alters properties by dissolution–precipitation reactions, producing new porosity and mechanically weak phases like chlorite and sericite [58]. These relationships provide the theoretical background for the variable nature of the behavior of the rocks studied.

The Shuangjiangkou granite specimens are examples of natural variability in crystalline rocks, with UCS values ranging from 116.39 MPa to 140.38 MPa (Figure 7). The differences in this variation are due to the individual microscopic and mesoscopic mineral, micro-discontinuity, and fabric orientations in each specimen. Mechanical anisotropy is introduced by the oriented distribution of minerals, especially phyllosilicates like mica, which leads to a variation in strength as a result of mineral fabric orientation to the direction of stress. This anisotropy, together with the intrinsic heterogeneity of the rock materials, leads to the differences in the observed mechanical behavior under similar loading conditions and the importance of the petrographic character in governing the mechanical response of rocks.

The Chengde anorthosite samples were tested for uniaxial compressive strength and exhibited high mechanical variation, with peak strengths ranging from 133.01 to 191.59 MPa (Figure 8). The Damiao anorthosite complex, mainly composed of plagioclase (85%–95%) with little pyroxene (5%–10%), Fe-Ti oxides, and apatite, has strength variations associated with mineralogical and textural heterogeneity [59,60]. The range of UCS observed indicates differences in the proportions of minerals and secondary alteration minerals, especially where dark anorthosite has been transformed into weaker, white forms. Mechanical behavior is greatly affected by textural properties, such as coarse-grained texture with a plagioclase crystal between 1 and 20 mm and megacrysts of 30 cm [10,61]. The grain interlocking and grain size distribution also contribute to the strength variability; thus, petrographic properties are indeed important in the control of the mechanical behavior of anorthosites.

The results of the uniaxial compressive strength test on specimens of Chengde mangerite varied between 165.62 and 239.94 MPa (Figure 9) due to inherent variability in the material properties. Variability in strength is mainly due to variations in the proportions of constituent minerals, including feldspar, quartz, and mafic minerals (such as biotite and amphibole). The difference in mechanical behavior is due to inherent strength differences between the minerals, where quartz and orthoclase are stronger than plagioclase and mafic constituents. Damiao complex mangerite is fine to coarse-grained and has phenocrysts, with variations in grain size distribution affecting its overall strength properties.

Table 1. Mechanical and physical properties of rock specimens.

Rocks	Sample ID	UCS (MPa)	UCS Mean	UCS SD	Absolute Axial Strain (%)	Volumetric Strain (%)	Water Absorption (%)	Porosity (%)	Specific Gravity
Shuangjiangkou Granite	SJK Gr A	145.68	127.53	15.07	0.46	0.23	0.25	0.75	3.03
	SJK Gr B	127.03			0.11	0.11	0.32	0.79	2.67
	SJK Gr C	116.39			0.88	1.38	0.38	0.88	2.40
	SJK Gr D	140.38			0.62	1.11	0.28	0.76	2.79
	SJK Gr E	109.17			0.55	0.97	0.42	0.92	1.88
Chengde Anorthosite	CD An A	133.01	167.81	23.45	0.47	0.59	0.44	1.20	2.67
	CD An C	190.97			0.66	0.77	0.11	0.31	2.76
	CD An D	175.82			0.67	1.15	0.12	0.33	2.75
	CD An E	146.83			0.66	1.19	0.18	0.47	2.68
	CD An F	191.59			0.73	0.91	0.07	0.20	2.90
	CD An G	168.65			0.53	1.06	0.12	0.33	2.72
Chengde Mangerite	CD Mgr A	212.01	205.12	23.87	0.65	0.78	0.12	0.35	2.77
	CD Mgr C	165.62			0.54	0.93	0.25	0.69	2.43
	CD Mgr E	202.59			0.46	0.67	0.22	0.60	2.70
	CD Mgr I	207.71			0.61	1.53	0.14	0.36	2.75
	CD Mgr K	203.84			0.72	1.08	0.15	0.42	2.72
	CD Mgr N	239.94			0.81	0.95	0.10	0.29	2.99

presents the strength and physical characteristics of the samples studied, revealing significant correlations between mechanical behavior and petrophysical properties. The physical properties (water absorption, porosity, specific gravity) are directly linked with void spaces and micro-fissures in the rock fabric [15,62]. The data from Shuangjiangkou granite show that the uniaxial compressive strength decreases with an increase in water absorption and apparent porosity (Figure 10). High water absorption means it has a greater ability to absorb water influx, and this increases the rate of weathering, causing a loss of strength. High porosity leads to low cohesion between mineral grains, resulting in low resistance to compressive stress. The positive correlation between specific gravity and UCS is also corroborated by the principles relating higher bulk density to decreased water absorption and porosity, thereby increasing compactness and strength. Bulk density is associated with mineralogical content, and it is inversely correlated with minerals such as albite and positively correlated with quartz and orthoclase [62].

The Damiao anorthosite specimens have a UCS range from 133.01 to 191.59 MPa, inverse correlations between UCS and water absorption (0.07% to 0.44%) and porosity (0.20% to 1.20%), and a positive relationship with specific gravity (2.67 to 2.90), as shown in Table 1. The increased value of water absorption and porosity means more void space, such as microcracks and alteration zones, that lowers the integrity of the material and cohesive strength between the grains. The entry of water increases the speed of weathering and consequently causes the loss of strength with time [17]. In contrast, the greater the specific gravity, the higher the density and compactness of rocks, and the less void space, leading to an increase in mechanical strength. Table 1 shows the UCS values of Chengde mangerite specimens to be between 165.62 and 239.94 MPa, with the inverse relationship between the strength and water absorption (0.10% to 0.25%) to porosity (0.29% to 0.69%) and positive correlation with specific gravity (2.43 to 2.99). The Damiao complex has undergone a process of hydrothermal alteration, resulting in changes to its mineral composition and textural alterations, as well as a direct influence on its physical properties. The rock structure contains planes of weakness, which are formed by cleavable minerals, especially feldspar, and twinning planes, cleavages, and microfractures [35,48]. The resulting UCS variability observed is a result of these physical properties and is dependent on the local mineralogy, texture, and alteration state within each specimen.

4.2. Petrographic and Mineralogical Characteristics

A polarizing microscope was used to analyze the thin sections to examine the following: (1) mineralogical characteristics; (2) microstructure/texture and weak planes; (3) alteration and weathering conditions; and (4) their role in influencing the physical and mechanical properties of the types of rocks. The average modal mineralogical composition of granite, anorthosite, and mangerite rocks is presented in Figure 11, whereas the corresponding range values are presented in

Table 2. Mineralogical composition (%) of granitic and mafic rocks.

Mineral	Granite	Anorthosite	Mangerite
Quartz (%)	25–27	-	10–15
K-feldspar (%)	40–45	-	1–4
Plagioclase (%)	-	88–92	40–50
Albite (%)	20–25	-	-
Perthite (%)	-	-	25–35
Orthopyroxene (%)	-	3–7	-
Muscovite (%)	5–7	-	-
Biotite (%)	2–3		-
Mafic minerals (%)	-	-	8–12
Tourmaline (%)	1	-	-
Apatite (%)	-	1–3	-
Fe-Ti oxides (%)	-	1–5	-

. The findings of each type of rock are explained in the subsequent sections.

4.2.1. Microstructural Features (Optical Microscopy)

(a)

Granite

The mineralogical composition of Shuangjiangkou granite is determined by petrographic analysis and consists of approximately 45% feldspar, 27% quartz, 10% mica, and minor amounts of tourmaline (Figure 12). Specifically, the sizes of quartz grains are hypidiomorphic and 2–5 mm wide, and feldspar grains are 1–5 mm in diameter. Quartz grains are anhedral to subhedral in morphology and possess undulatory extinction, causing strain accumulation due to prior tectonic stress (Figure 12A–F). Furthermore, quartz crystals exhibit microfractures and grain boundary migration, which serve as a preferential failure path during loading, resulting in increased porosity and a reduced UCS. Alkali feldspar (microcline, orthoclase) and plagioclase (albite) (Figure 12A,B,D) exhibit some degree of alteration. The microcline has typical cross-hatched twinning, as well as sericitization of some grains along cleavage planes. The alteration process raises the degree of microporosity and water absorption capacity, which is why the absorption of water and porosity is high in comparison to the other types of rocks (Table 1). Further, the feldspars have perfect cleavage in two directions, creating natural planes of weakness that allow crack propagation under stress.

Biotite and muscovite mica crystals are found in elongated, tabular, and rod-shaped flakes (Figure 12A,B,D,E). Biotite exhibits partial chloritization, indicative of retrograde metamorphism, which weakens grain boundaries (Figure 12A). The anisotropic mechanical properties are formed by the preferred orientation of mica flakes, which are weaker along the foliation. The high mica content (8.5%) directly affects water absorption because of its hydrophilic character and expandable crystal structure. Moreover, the biotite–muscovite intergrowths develop areas of mechanical inhomogeneity, where the differences hinder stress transfer in the elastic properties of the adjacent minerals (Figure 12D). Tourmaline crystals are prismatic euhedral grains with typical triangular cross-sections (Figure 12A,C). Notably, Figure 12C shows extensive fracturing of tourmaline grains, indicating brittle deformation even though tourmaline has a relatively high hardness (7–7.5 on Mohs scale). These existing microfractures act as stress concentrators, causing failure at lower applied stresses than predicted by intact mineral strength.

Overall, granite exhibits higher mica composition, distinct microstructural features (weak planes, sheet mica, fractures), and chemical weathering that significantly influence its physical properties and mechanical behavior. The role of these properties has previously been established [63,64]. These petrographic alterations on the rock fabric are attributed to significantly high porosity, low specific gravity, and low mean UCS of the granite (Table 1).

(b)

Anorthosite

The anorthosite rock samples investigated from North China Craton are creamy white in hand specimen, brittle, with a network of fine microfractures (Figure 13A). Generally, the rock is medium- to coarse-grained and mainly consists of plagioclase phenocrysts, which comprise over 90% of the modal content (Figure 13B,D,F,G). The main mafic mineral is orthopyroxene (5%), with magnetite, ilmenite, and apatite (2%) as the accessory minerals. The plagioclase is altered; however, the typical polysynthetic twinning is retained. Figure 13B shows a high degree of fracturing of the plagioclase crystals around ore minerals forming inter-connective networks, which probably enhances higher water absorption and porosity of weaker specimens. The alteration minerals are clinozoisite and sericite (Figure 13C), which develop along the cleavage planes and grain boundaries. This phyllosilicate alteration is significant because sericite has a platy morphology and a perfect basal cleavage, so it forms planes of weakness that weaken the cohesion between plagioclase grains. Some plagioclase crystals show kinks in their twin lamellae, fractured by micro-fissures or granulated at the edges (Figure 13D,F,G), indicating recrystallization during hydrothermal alteration. Figure 13D shows a massive micro-fracturing within plagioclase crystals, and the fractures are preferentially oriented in cleavage planes. The intracrystalline discontinuities become stress concentrators under loading and induce the crack propagation at lower applied stresses than would be needed in the unaltered material.

Figure 13E shows the typical occurrence of pyroxene with ore minerals, where the mechanical contrast of these phases results in a differential stress state. Almost all the primary pyroxenes are altered to fine-grained aggregates of uralite, magnetite, and chlorite, and the primary textures are usually retained. Chlorite alteration products (Figure 13F) are weaker than the original pyroxene and form areas of weak mechanical strength within the rock matrix. Figure 13G shows another essential textural aspect where the margins of the plagioclase crystals are altered. These reaction rims are regions of the most intense fluid–rock interaction; correspondingly, they exhibit low grain boundary cohesion. The aggregation of alteration products around grain boundaries forms a lattice of possible failure planes that allows the coalescence of cracks during compression.

The integration of petrographic observations with mechanical test results illustrates that the strength variability within the Chengde anorthosite is mainly associated with the degree and mode of hydrothermal alteration and microstructures. Specimens with lower UCS values consistently show more extensive sericite–chlorite alteration, higher microfracture density, and greater grain boundary alteration. In contrast, stronger specimens retain more of their primary igneous texture with minimal alteration. In comparison to granite, anorthosite exhibits fewer deformational features (Figure 12 and Figure 13), which is also reflected in the physical and mechanical test results (Table 1).

(c)

Mangerite

Mangerite samples from the Damiao anorthosite complex are brown-dark gray in hand specimen (Figure 14A), with a heterogeneous texture characterized by visible feldspar phenocrysts and relatively low alteration and deformation features. The rock is composed of subhedral perthite (20%–40%), plagioclase (40%–50%), interstitial quartz (5%–20%), and minor K-feldspar (2%–3%), with accessory minerals variable including clinopyroxene (augite), orthopyroxene (hypersthene), green hornblende, biotite flakes, Fe-Ti oxides (magnetite–ilmenite intergrowths), apatite, and sporadic olivine rimmed by reaction coronas.

The typical perthitic texture of micro-fracturing through perthite and plagioclase crystals is depicted in Figure 14B. These existing cracks, in addition to secondary calcite vein filling (Figure 14C), create planes of weakness that allow propagation of cracks under compressive loading. The oxidized regions observed along the grain boundaries are indicators of the fluid percolation paths that may destabilize the intergranular cohesion under loading.

The degree of feldspar alteration is limited to some samples (Figure 14C,D). Notably, some plagioclase and K-feldspar grains show pervasive alteration to sericite, suggesting late-stage hydrothermal activity. Figure 14E reveals the mineralogical heterogeneity characteristic of mangerite, with hornblende, quartz, and variably altered feldspars creating a complex mechanical framework. The quartz grains, being mechanically stronger and more resistant to alteration, form competent domains within the rock. The samples are pervasively fractured; at places, these fractures are typically filled by secondary calcite and minor chlorite minerals (Figure 14F). These healed fractures provide some cohesion, also noticed in previous studies [65]. The higher mechanical properties (Table 1) observed in the mangerite can be attributed to several mineralogical and textural factors. First is the higher %age (>95%) of hard minerals (quartz, perthite, plagioclase, k-feldspar, and mafic minerals). Second is the absence of weak sheet minerals (mica) and secondary minerals such as sericite, chlorite and calcite. Third, there is a low proportion of fracturing and alteration of the fabric. This is reflected in the mangerite’s significantly higher physical and mechanical properties.

4.2.2. Mineralogical Composition and Texture (SEM, EDS)

The mineralogical composition and microstructural texture of the rock material were studied using Scanning Electron Microscopy (SEM) and Energy-Dispersive X-ray Spectroscopy (EDS).

(a)

Granite

Back-scattered electron (BSE) imaging showed well-interlocked crystalline fabric mainly composed of quartz and feldspar, with accessory biotite and mica in granite. Internal pore structures (about 500 μm in diameter) and discrete microcracks (related mainly to grain boundaries) were also observed in the images (Figure 15). At spectrum point 15–18, quartz grains, SEM-EDS analysis indicated a major composition of O (70.36%), Si (29.60%), and trace Zr (0.14%). The presence of zirconium in quartz grains is remarkably rare, as the Zr⁴⁺ ions (0.72 Å) rarely substitute Si⁴⁺ (0.40 Å) because of size incompatibility. This replacement is likely indicative of high-temperature (>600 °C), highly evolved, zirconium-rich magmatic environments associated with late-stage granite differentiation and potential rare-metal mineralization. Trace zirconium may locally enhance the stiffness and strength of quartz domains, thereby increasing the load-bearing capacity. Images of BSE indicated that fractures tend to initiate and propagate along mineral surfaces, compared to the grain interiors, especially in the feldspar-rich areas. The pattern of localized damage enables the redistribution of stress in structurally competent quartz domains and sustains overall mechanical integrity. The grain interlocking texture that is essential to the strength of granite was maintained to a large extent in the samples that had evidence of geochemical alteration. The overall microcrack density (0.3–0.5 cracks/mm²) lies within the normal range of intact granite [63]. Thus, the decrease in mechanical performance may be more related to mineral alteration than to physical damage.

At spectrums 33–36, K-feldspar is a light gray phase with an elemental composition of O: 65.39%, Na: 1.28%, Al: 7.28%, Si: 20.94%, Cl: 0.45%, and K: 5.64%, as shown in Figure 16, which highlights the BSE contrast between K-feldspar and surrounding minerals. The presence of 1.28% Na indicates there were cryptoperthitic intergrowths by exsolution in the cooling process, producing internal planes of weakness to facilitate fracture growth [66]. The aberrant content of Cl (0.45%) points to the hydrothermal alteration at temperatures above 400 °C that adds to the hydrolytic weakening along the cleavage planes and contributes to the observed variability of the UCS (109.17–145.68 MPa) [67]. BSE imaging indicates quartz grain fracturing and intergranular merging with feldspar, probably caused by high-pressure, high-temperature metamorphism. These contacts foster local weakening. The feldspar and quartz have different thermal expansion, which creates boundary stresses of 20–50 MPa to promote crack initiation. Moreover, medium Na-K feldspar weakens granite by ~15% relative to K-feldspar endmembers, and every 10% increment in the content of feldspar can lower UCS by 15–25 MPa due to increased cleavage development.

(b)

Anorthosite

Back-scattered electron (BSE) imaging showed important mineralogical features that determine the mechanical behavior of anorthosite. Figure 17 shows the presence of albite (NaAlSi₃O₈) in existing fractures, with EDS analysis showing an elemental composition of O: 63.93%, Na: 6.51%, Al: 8.55%, Si: 20.63%, and Ca: 0.58%. Mechanical anisotropy is also induced by the presence of albite in the fractures, which has perfect [001] cleavage and a lower hardness as compared to the surrounding plagioclase matrix. These albite-filled planes are directions of preferential failure of rock subjected to uniaxial loading and decrease the overall strength of the rock fabric. This mineralogical heterogeneity is a major source of variation in UCS, with variations exceeding 58 MPa among samples of similar composition, which emphasizes the importance of the distribution of minerals in the microstructure in determining the strength properties of anorthosites.

Further analysis of BSE in Figure 18 revealed deformed orthopyroxene grains with an elemental composition of O: 67.69%, Mg: 6.43%, Al: 9.67%, Si: 9.77%, and Fe: 6.43%. The decreased BSE contrast reveals pseudomorphic replacement resulting from hydrous alteration. Fresh orthopyroxene normally increases both the strength and the cohesiveness of mafic rocks; however, when altered, it loses more strength and becomes less cohesive phases with reduced elastic moduli. These deformations induce mechanical heterogeneity in the rock, which lowers the strength of the rock. Fractures are likely to originate and extend within these modified areas during loading, and as such, an increase in uniaxial compressive strength variability has been observed between anorthosite specimens.

BSE imaging (Figure 19) showed a major fracture zone with contrasting compositions of mineral phases. Titanium oxide (probably ilmenite) and an elemental composition of O: 75.20%, Al: 0.65%, Si: 1.68%, Ca: 0.59%, and Ti: 21.89% were determined as bright phases. Adjacent plagioclase had medium BSE contrast. Mechanical heterogeneities are caused by the concentration of compositionally disparate minerals near fractures, forming susceptible points of stress concentration that lead to the initiation and propagation of cracks. This inhomogeneous filling up along existing discontinuities reduces the structural integrity of the rock and leads to the variance in the uniaxial compressive strength across the anorthosite samples.

The SEM-EDS analysis shows that the mechanical response of Chengde anorthosite is highly dependent on its dense crystalline structure and localized heterogeneity of minerals. Mechanical discontinuities are caused by fracture-filling albite and altered orthopyroxene zones despite the low porosity. Specimens that are not significantly altered and have fewer filled fractures have UCS values close to 192 MPa, whereas those with excessive microstructural defects decrease to 133 MPa. The difference in mineralogical and textural features, caused by ~44% variation, highlights the importance of mineralogy and textural features of stress redistribution and crack propagation. These results indicate the importance of microstructural characterization in the variation in the strength of a nominally homogeneous lithology and the accuracy of rock mass quality evaluations.

(c)

Mangerite

The BSE imaging and EDS analyses of the mangerite samples indicated considerable feldspar heterogeneity along fracture zones. Spectrum analysis at points 23 and 24 revealed that plagioclase feldspar was the predominant phase with a composition of O: 67.74%, Al: 11.03%, Si: 13.24%, Ca: 6.58%, and Fe: 1.42% (Figure 20), indicating intermediate composition (An30–40) related to mangeritic origins. Comparatively, points 25 and 26 of the spectra revealed albite (NaAlSi₃O₈) having O: 64.04%, Na: 7.39%, Al: 7.67%, and Si: 20.92%, as a secondary phase occupying the fracture. The almost pure albite indicates late-stage hydrothermal alteration/recrystallization, coming after primary magmatic crystallization. Mechanically, this feldspar association presents structural heterogeneity. The characteristic twin cleavages of plagioclase produce orthogonal planes of weakness, and the albite presence in fractures also leads to mechanical anisotropy because of the modification of bonding strength between minerals and matrix. The difference in uniaxial compressive strength (UCS) as high as 74.32 MPa between different specimens is associated with feldspar orientation, fracture density, and alteration intensity. Despite its high chemical stability, albite, as a fracture-filling mineral, can cause weaker bonding interfaces, thus generating favorable crack initiation points when subject to compressive stress. Such microstructural characteristics have a direct impact on the failure mode and strength variability of mangerite.

The low porosity and low water absorption of Chengde mangerite show a high degree of crystalline interlocking despite the fracture-filling minerals. Analyses of BSE and EDS validate that these cracks are sealed with secondary feldspar phases, which lowers permeability and ensures the integrity of the structure. The presence of a strong correlation between specific gravity and UCS shows a strong compositional control of strength, whereby the growth of orthopyroxene content leads to improved mechanical performance through the stiffer and more fracture-resistant nature of pyroxenes compared with feldspars. The combination of these microstructural and compositional factors can be assumed to be responsible for the resultant variation in strength among mangerite specimens.

5. Discussion

5.1. Petrography Control on Physico-Mechanical Properties

This study enhances our knowledge of rock mechanical behavior by providing a complete quantitative correlation between petrographic characteristics and failure modes in granitic and mafic rocks, unveiling critical threshold values with important implications for engineering practice. The experimental data reveal that the uniaxial compressive strength (UCS) varies considerably within the same type of rock and between different rocks as well, with granite samples having 109.17–145.68 MPa, anorthosite 133.01–191.59 MPa, and mangerite 165.62–239.94 MPa (Table 1). These differences are directly related to the mineral composition, texture, and microstructural features of the materials, as has been previously observed in the literature, where a strong correlation has been found between petrographic properties and mechanical behavior [64].

While past literature has revealed overall relationships between mineralogy and strength, this study establishes the predictive potential of trace element signatures in defining mechanical weakness. The mineralogical composition analysis explains fundamental differences between the three rock types, which explain their distinct mechanical behaviors. The granite samples are characterized by 42.5% K-feldspar, 26% quartz, 22.5% albite, some mica, and accessory minerals (Table 2). The relatively high quartz content helps with the brittle nature of failure of the rock because of the lack of cleavage and high inherent strength (Mohs hardness 7) of quartz, as indicated by Přikryl [68], who showed that quartz content has a positive correlation with rock strength. But the presence of feldspar minerals, which form more than 65% of the rock volume, provides planes of weakness by their perfect cleavage in two directions [69]. Another new finding of this study is the characterization of abnormal trace element compositions by EDS analysis, which reveals 0.14% Zr in quartz and 0.45% Cl in K-feldspar as diagnostic features of post-magmatic hydrothermal alteration that affects the original mineral chemistry and weakens grain boundaries [70]. Anorthosite is composed mostly of plagioclase feldspar (>90%), minor orthopyroxene (5%), and accessory Fe-Ti oxides (Table 2). This near-monomineralic composition makes the mechanical framework more homogeneous. However, much of it has been altered by sericitization and chloritization on cleavage planes, which has a major effect on strength [71]. BSE images showed albite-filled fractures that had compositions of Na: 6.51%, Al: 8.55%, and Si: 20.63%, which formed planes of mechanical weakness because of the perfect [001] cleavage and relatively low hardness of albite in comparison with the plagioclase matrix. Mangerite shows intermediate properties containing 45% plagioclase, 30% perthite, 12.5% quartz, and 10% mafic minerals (Table 2). The more varied mineral assemblage renders it more mechanically heterogeneous, although the increased amount of mechanically competent minerals (quartz and fresh pyroxene) leads to better strength values [72].

The size distribution of grain, as well as grain texture, has a significant impact on mechanical behavior, as widely reported in the rock mechanics literature. Granite samples are medium to finely grained in texture and have a grain size of 0.20 mm to 4.75 mm, in which smaller grain sizes are typically associated with higher strength values according to the Hall–Petch effect [73]. However, grain boundaries in altered samples are weak when they are filled with secondary minerals such as sericite and chlorite [74]. There is a strong textural control of strength in the anorthosite samples, with crystal sizes of plagioclase of 1–20 mm and rare megacrysts up to 30 cm. This coarse-grained structure features reduced grain boundaries that deflect cracks while maintaining strong crystalline interlocking, ensuring that bulk strength remains high [28]. Damage in plagioclase crystals in the form of kinked twin lamellae and marginal granulation is evidence of deformation-induced damage that forms stress concentrators, as observed by the micromechanical analysis of Fredrich [69]. Experimental data show high correlations between physical properties and mechanical strength (Figure 21 and Figure 22a–c). The negative correlation between porosity and UCS is indicative of the basic effect of void space in decreasing effective bearing area and in causing stress concentration [75]. The average porosity of granite specimens is the highest (0.82%) of the three types of rocks, consistent with the low average strength of granite. Feldspar dissolution and mica alteration lead to the formation of inter-connected pore networks, allowing for additional weathering [76].

Water absorption acts as a proxy for total porosity and pore connectivity, exhibiting the strongest negative correlation with UCS (r = −0.82), which is consistent with prior studies [22,75]. Water penetration into the rock structure indicates that pore networks and microfractures are connected and affect the mechanical integrity. Another important result of the research is the determination of a critical threshold value: specimens of anorthosite and mangerite with low water absorption levels (below 0.15%) all exhibit UCS values greater than 180 MPa, whereas specimens with higher absorption values show significantly decreased strength. This threshold provides a practical engineering tool for rapid rock quality assessment and contributes to predictive rock mechanics, as implied by the threshold values identified by Tugrul and Zarif [64]. This alteration process is observed in a high negative correlation between the UCS values and water absorption at all rock types (Figure 21). Specific gravity also proves to be a good predictor of rock quality and is highly positively correlated with UCS, as previously found by Bell [77] and Kahraman [75]. Greater specific gravity values point to more compact mineral grain packing, fewer spaces, and, in general, less alteration. This relationship is the most intense in granite, where specific gravity varies between 1.88 and 3.03 depending on levels of weathering and mineral change. The specific gravity of certain granite specimens is anomalously low and is associated with widespread feldspar kaolinization and mica vermiculitization, which lowers bulk density greatly and generates mechanically weak phases [70].

5.2. Failure Mode Analysis of Specimens

This study proposes a quantitative systematization of failure modes based on crack density distribution, establishing a new framework for comprehending the influence of petrographic variations on the deformation processes of crystalline rocks. The post-failure analysis of cylindrical specimens subjected to uniaxial compressive strength testing has shown characteristic patterns of failure that strongly correlate with the type of rock and the underlying petrographic properties. The failure modes observed are of great importance in understanding how mineralogical composition and microstructural characteristics govern the mechanical behavior and deformation processes. Frictional resistance between the specimen ends and the loading platens interferes with the failure modes; to prevent this, petroleum jelly (Vaseline) was used as a lubricant to reduce end friction during loading. It is generally applied in rock mechanics testing to control the interpretation of failure mechanisms more accurately and to simulate frictionless conditions at boundaries [78,79]. Friction between the specimen and the platen can prevent lateral deformation and induce artificial shear or barreling; therefore, minimizing friction between the specimen and the platen is essential to observe true split-axial or shear failures. The experimental results, which display failure modes with appropriate lubrication and a specimen geometry (height-to-diameter ratio of 2:1), are representative of the intrinsic mechanical behavior of the rock specimens tested and provide reliable insights into how petrographic and mineralogical variations influence failure processes in granitic and mafic rock types. This work contrasts with previous qualitative descriptions of failure modes and defines quantitative relationships based on mineralogical composition and crack development patterns. The failure modes of Shuangjiangkou granite specimens were mainly shear-dominated, with well-developed diagonal shear planes with a trend of 30–45° with the vertical axis of the loading (Figure 23a). The prevalence of shear failure mechanisms is directly connected to the heterogeneous mineralogical composition and microstructural features in granite. The anisotropic mechanical properties are strong and controlled by the presence of 10% mica content, which has a platy morphology and perfect basal cleavage in the form of biotite and muscovite minerals that act as the planes of preferred failure that enable shear deformation. Optimal mica flake orientation creates planes of weakness that are normal to the foliation and drive the formation of shear bands during compressional deformation. Moreover, widespread feldspar alteration (45% of the total composition) to sericitization along cleavage planes considerably enhances the microporosity. It decreases the intergranular cohesion, presenting an optimum environment for the propagation of shear cracks. The undulatory extinction observed in the quartz grains (27% composition) indicates past tectonic strain accumulation, where pre-existing microfractures and grain boundary migration phenomena act as stress concentrators, inducing failure at lower applied stress values and resulting in conjugate shear patterns.

Anorthosite specimens (Figure 23b) showed quite different failure patterns, with predominant tensile fracturing and several vertical cracks forming typical splitting patterns. The tensile-dominated failure mode is indicative of the homogeneous mineralogical composition of the rock, characterized by more than 90% plagioclase phenocrysts, which provides a more even stress distribution than the heterogeneous granite matrix. However, extensive plagioclase alteration to clinozoisite and sericite along cleavage planes, kinked twin lamellae, and intracrystalline microfractures generate networks of possible failure surfaces that allow tensile cracks to coalesce. The transformation of orthopyroxene (5%) into fine-grained clusters of uralite, magnetite, and chlorite creates areas of mechanical weakness with lower shear strength than primary pyroxene, encouraging preferential crack propagation along altered boundaries.

Mangerite specimens show the most extensive fractures and dense networks of large tensile cracks, producing substantial fragmentation of the specimen (Figure 23c). This disastrous brittle failure is directly related to the excellent petrographic features of mangerite, such as the presence of >95% hard minerals (quartz, perthite, plagioclase, K-feldspar, and mafic minerals) and a significant lack of weak sheet minerals. The greater proportion of mechanically competent minerals, especially those with 5%–20% quartz content, forms resistant domains, establishing a highly interlocked crystalline structure with elevated elastic moduli. The perthitic texture and the low degree of alteration make the intergranular bonding very strong, and the low proportion of micro-fracturing leads to minimal stress concentration points. As a result, a large amount of stored elastic energy is dissipated in a short time due to numerous tensile fractures, which is why the catastrophic mode of failure was observed.

Visual analysis reveals distinct quantitative variations in crack density resulting from petrographic differences. Granite samples formed 1–3 major shear cracks as a result of gradual failure along mica-rich weaker planes, anorthosite samples had 4–8 primary tensile cracks induced by the mica alteration effect, and mangerite samples had 6–12 tensile cracks formed by rapid crack propagation through the strong mineral network after the elastic limit was surpassed.

This study shows that the micro-scale mineralogical characteristics have direct control on the macro-scale mechanical behavior of crystalline rocks. The discovery of post-magmatic alterations, such as zirconium in quartz and enrichment of chlorine in feldspar, reveals new mechanisms by which geochemical changes can systematically weaken grain boundaries. The contribution of albite-filled fractures to the tensile splitting of anorthosite and the catastrophic fragmentation behavior of mangerite, resulting from a highly interlocked mineral framework, presents new perspectives on failure mechanisms. These results indicate that petrographic observations have the potential to systematically predict mechanical behavior and link microstructural analysis with rock mechanics through quantitative relationships applicable to engineering evaluations and material classification systems.

6. Conclusions

This comprehensive study developed quantitative correlations between mineralogical and petrographic characteristics and the mechanical behavior of granitic and mafic rocks in the Shuangjiangkou and Damiao complexes. The integrated methodology combining detailed petrographic analysis using optical microscopy and SEM-EDS with systematic geotechnical characterization has yielded significant insights. The main findings are summarized as follows:

The mineralogical composition has a primary influence on mechanical performance, and the granite samples that had 45% feldspar, 27% quartz, and 10% mica had a mean uniaxial compressive strength of 127.53 ± 15.07 MPa, and anorthosite with >90% plagioclase had 167.81 ± 23.45 MPa. Mangerite, with >95% hard minerals, attained the highest strength of 205.12 ± 23.87 MPa, showing the fundamental control of mineral assemblage on mechanical behavior.

Strong inverse correlations exist between mechanical strength and physical properties, with water absorption showing strong negative correlations with UCS for granite (R² = 0.98), anorthosite (R² = 0.74), and mangerite (R² = 0.76), while porosity shows similar strong relationships for granite (R² = 0.93), anorthosite (R² = 0.72), and mangerite (R² = 0.75). Specific gravity is positively correlated with strength, particularly strong for mangerite (R² = 0.99) and moderate for granite (R² = 0.89) and anorthosite (R² = 0.66).

Critical petrographic features directly control mechanical behavior through multiple mechanisms. High mica content creates anisotropic properties in granite; feldspar alteration increases microporosity, and plagioclase micro-fracturing creates weak networks in anorthosite, while mangerite’s hard mineral composition ensures exceptional mechanical properties.

Microstructural characteristics directly influence failure mechanisms. SEM-EDS analysis shows granite exhibits preferential fracture propagation along mineral interfaces, anorthosite shows albite-filled fractures introducing mechanical anisotropy, while mangerite displays excellent crystalline interlocking with secondary phases sealing fractures and maintaining structural integrity.

Distinct failure patterns correlate with petrographic characteristics; granite exhibits shear-dominated failure (30–45° diagonal planes) due to mica-induced anisotropy and feldspar alteration. Anorthosite exhibits tensile fracturing with vertical splitting, resulting from a homogeneous plagioclase composition (>90%) that is compromised by alteration networks. Mangerite shows catastrophic brittle failure (6–12 tensile fractures) through rapid elastic energy release.

The findings provide quantitative frameworks that link petrographic characteristics to engineering behavior, enabling the predictive assessment of rock behavior based on petrographic analysis and contributing to improved engineering geological practice.