The Effect of Volcanic Stone and Metakaolin on the Compressive Properties of Ultrahigh-Performance Concrete Cubes

Abstract

Over the past few decades, ultrahigh-performance concrete (UHPC) has been widely studied and applied because of its outstanding mechanical properties, such as its high strength and notable durability. However, because of its high cost and easy shrinkage cracking during early pouring in mass concrete construction, to reduce the cost of UHPC and reduce the cracks caused by early pouring, volcanic stone was used as a new type of UHPC coarse aggregate, while metakaolin (MK) was added to the system at the same time, and then two parameters, namely the volcanic rock particle size group and the MK dispersion ratio, were set. Scanning electron microscopy (SEM), X-ray diffraction (XRD) and thermogravimetric (TG) microanalysis methods were used to reveal the influence of changes in the material microstructure, phase composition, material composition and crystallinity of the mineral composition on the compressive properties of the UHPC cubes. The results show that the mechanical “lock-in effect” of the structure formed by the volcanic rock holes and mortar can effectively improve the shear resistance of the UHPC–volcanic rock interface, and the compressive strength of the UHPC cubes increases with the volcanic stone’s particle size. When the MK dispersion ratio is less than 4%, the cube compressive strength of the UHPC and the contents of CaCO₃ crystals, C-S-H gel and travertine in the UHPC increase with an increasing MK dispersion ratio. At an age of 28 days, compared with the addition of 1% MK, the addition of 4% MK increases the production of C-S-H gel and travertine in the UHPC matrix by 24.82%. When the MK dispersion ratio is 4%, the crystallinity values of the C-S-H gel, travertine and limestone in the UHPC are greater. Adding MK at a 4% dispersion ratio can promote the crystallization of limestone into a large amount of calcite, which can increase the strength of UHPC. On the one hand, the addition of volcanic coarse aggregate results in the retention of more free water and bound water; on the other hand, it also makes it difficult to crystallize CaCO₃. The combined action of MK at a 4% dispersion ratio and volcanic rock significantly inhibits CaCO₃ crystallization.

1. Introduction

Ultrahigh-performance concrete (UHPC) is a durable and high-performance concrete material with excellent properties, such as high strength and durability [1,2]. Since its emergence more than 20 years ago, UHPC has attracted the attention of civil engineering circles at home and abroad. While UHPC achieves ultrahigh performance, the reality of its high production cost cannot be ignored. The cost of its raw materials is generally above 4000 RMB/m³, with the highest reaching 10,000 RMB/m³ [3]. In addition, due to its low water–binder ratio and large amount of self-shrinkage, UHPC is prone to shrinkage cracking during the early casting process [4]. Early cracking does not close, which not only reduces the compressive strength of UHPC in the later stage but also affects the overall quality of the project. High cost and high shrinkage seriously restrict the promotion and applications of UHPC in mass concrete engineering [5,6,7,8].

In response to the above problems, scholars at home and abroad have proposed the formation of coarse aggregate–UHPC (CA-UHPC) containing coarse aggregate by adding natural coarse aggregate to UHPC to replace part of the active powder without significantly reducing the excellent mechanical properties of UHPC. CA-UPC is generally made by replacing UHPC components with basalt gravel of a certain particle size. For example, Shi et al. reported that the compressive strength of CA-UHPC is not much different from that of UHPC without coarse aggregates [9,10,11,12], Wille et al. reported that coarse aggregates with larger particle sizes can improve the compressive strength of UHPC to a certain extent [2] and the compressive strength of CA-UHPC tends to first increase and then decrease with an increasing coarse aggregate-to-binder ratio [13,14].

Studies have found that metakaolin (MK) not only improves the early strength of concrete but also enhances its durability and resistance to chemical attack. Metakaolin reacts pozzolanically to produce additional calcium silicate hydrate (C-S-H), filling the pores in the concrete, thus increasing its density and crack resistance [15,16]. Additionally, metakaolin exhibits water-reducing properties, which can lower the water demand in concrete and improve its workability [17]. The combined use of volcanic ash and metakaolin in concrete demonstrates significant synergistic effects. Research indicates that the combination of these two materials can further enhance the compressive strength, tensile strength and durability of concrete [18,19]. Studies have also shown that the specific surface area of MK has a significant effect on cement-based materials [20], and the filling effect and active reaction of MK can effectively improve the early mechanical properties of concrete [21,22], improve the pore structure of cement and improve its self-shrinking ability; at the same time, the addition of MK can fully improve the strength and toughening ability of steel fibers [12]. Microscopically, the amount of chemical bonding water in the gel in the UHPC matrix increases with an increasing MK content, while the amount of Ca(OH)₂ decreases continuously [23,24]. When MK is used to replace 10~15% of the cement base, the compressive performance of UHPC can reach the best level [25,26,27].

Scholars have effectively improved the early mechanical properties of concrete by carrying out structural improvement and material modification of cement-based materials, especially for the placement of coarse aggregates in UHPC, which has led to some beneficial research results. However, research on the effect of MK on the properties of cement-based materials has been concentrated on non-high-strength concrete materials, and research on its impact on UHPC’s mechanical properties is still relatively rare; moreover, related research on the impact of MK on the mechanical properties of CA-UHPC is rare. In addition, basalt gravel is generally used as a coarse aggregate in UHPC, while other coarse aggregate substitutes have rarely been studied. Thus, the limitations of the related research are obvious.

In this paper, by using volcanic rocks as a new type of coarse aggregate for UHPC, the goal was not only to maintain the compressive strength of UHPC but also to significantly reduce the construction cost of UHPC projects, resulting in the development of a new type of high-strength engineering material. The conclusions of this paper provide reliable data support for the UHPC theoretical framework and its application in large-scale green CA-UHPC engineering projects.

2. Experimental Program

2.1. Material Selection

As shown in Figure 1, the cement is ordinary Portland cement with a strength grade of P·O42.5, produced by the Liaoning Daying Cement Plant in China, and its loss on ignition (LOI) is 3%.The silica fume was produced by Lingshou Yunda Mineral Products Co., Ltd. (Huainan, China), with a particle size of 2.5 μm, and its LOI is 0.5%. MK was produced by Inner Mongolia Baotou Steel Hefa Rare Earth Co., Ltd. (Baotou, China), with a particle size of 3.75 μm. River sand with a particle size of 0.075~2.36 mm from certain tailing impoundments in Liaoning Province was adopted. Volcanic stone with a particle size of 9.5~31.5 mm was obtained from Lingshou Weifeng Mineral Products Co., Ltd. (Huainan, China), which has a crushing value of 24.36%. The steel fibers used were steel fibers produced by Tengzhou Sida Shimisi Metal Products Co., Ltd. (80% end-hook-type and 20% straight-wire-type, length of 13 mm, length–diameter ratio of 65, tensile strength of 2850 MPa, Tengzhou, China), and the water-reducing agent used was a polyhydroxy acid superplasticizer mother liquor produced by Weihe Technology Admixture Factory (water-reducing rate of 40%, Lianyungang, China). All the raw materials were sourced from China. The chemical compositions of the cement and kaolin are shown in

Table 1. Chemical composition of cement and MK (w/%).

Raw Material	SiO2	Al2O3	Fe2O3	CaO	MgO	Na2O	SO3	K2O	TiO2	The Others
Cement	23.22	4.51	2.33	60.47	1.16	0.87	4.22	1.23	-	1.99
MK	51.92	44.52	0.42	-	-	-	-	-	1.23	1.91

.

2.2. Mix Proportion Design

The dispersion ratio of MK refers to the proportion of MK in cementitious materials. Fly ash is not used in this study because it is also a cementitious material and would affect the analysis of MK’s impact on cement performance. Therefore, the MK proportion is expressed as a percentage total mass of cement and silica fume (1–4%). The factor level of the experimental design is shown in

Table 2. Parameters and levels.

Levels	Parameters
	A: MK Dispersion Ratio/%	B: Volcanic Stone Granular Formation/mm
1	1	9.5–16
2	2	16–19
3	3	19–31.5
4	4	-

. The MK dispersion ratio and the volcanic stone particle size group were used as two variable factors, A and B. The A factor level was 4, and the B factor level was 3. At the same time, a UHPC mix ratio design table was established, with a total of 12 groups, as shown in

Table 3. Mix proportion of UHPC.

No.	Mass Fraction/%							Water–Binder Ratio	Particle Size of Volcanic Stone/mm
	Cement	Silica Fume	MK	Volcanic Stone	River Sand	Steel Fiber	Water-Reducing Agent
U	35.46	8.00	0	0	40.0	6.28	0.87	0.18	0
G1	35.46	8.00	0.43	0	40.0	6.28	0.87	0.18	0
G2	35.46	8.00	0.86	0	40.0	6.28	0.87	0.18	0
G3	35.46	8.00	1.30	0	40.0	6.28	0.87	0.18	0
G4	35.46	8.00	1.73	0	40.0	6.28	0.87	0.18	0
H1	29.79	6.72	0	16	33.6	5.28	0.73	0.18	9.5–16
H2	29.79	6.72	0	16	33.6	5.28	0.73	0.18	16–19
H3	29.79	6.72	0	16	33.6	5.28	0.73	0.18	19–31.5
GH1	29.79	6.72	0.36	16	33.6	5.28	0.73	0.18	9.5–16
GH2	29.79	6.72	0.73	16	33.6	5.28	0.73	0.18	9.5–16
GH3	29.79	6.72	1.09	16	33.6	5.28	0.73	0.18	9.5–16
GH4	29.79	6.72	1.46	16	33.6	5.28	0.73	0.18	9.5–16

Notes: U stands for UHPC; G represents UHPC doped with MK (GA, a number represents the horizontal factor of factor A in Table 2); H represents UHPC mixed with volcanic stone (Hb, B represents the horizontal factor of factor B in Table 2); GH represents UHPC mixed with MK and volcanic stone (GHC, C represents the horizontal factor of factor A in Table 2).

.

2.3. Test Specimens

With 6 specimens as a group, a total of 72 cube compressive standard specimens were used to carry out the experimental research. By calculating the weights of the required materials, the MK and water were mixed according to the mixing ratio and stirred evenly to make the MK slurry, and the volcanic stone was soaked in the MK slurry for 12 h and then removed. Then, the weighed cementing material (cement, silica fume), river sand, volcanic stone, steel fiber, water-reducing agent and MK slurry were poured into the mixer, wetted by the cement slurry in turn. Each time a material was poured, it was stirred for 30 s until it was uniform. After all the raw materials were added, the mixture was stirred again for 2 to 3 min and then quickly poured into a 150 mm × 150 mm × 150 mm trial mold. After pouring, the test specimen was placed in a normal-humidity environment of 20 ± 1 °C for 48 h and then demolded. After demolding, the test specimens were put into a standard curing box with a temperature of 20 ± 1 °C and a humidity greater than 95% and cured for 3 days, 7 days, 14 days or 28 days. The compressive strength of the cubes was tested at the corresponding curing age, and XRD and TG microscopic tests were carried out at the corresponding curing age.

2.4. Microscopic Test Method

To observe the microscopic morphology of the materials and the specimens, according to the scanning electron microscopy (SEM) sampling requirements, samples of 10 mm × 10 mm × 10 mm specimens were collected after destruction. The samples were fixed, rinsed, dehydrated, replaced, dried and coated. The processed samples were scanned by SEM point scanning using a Shimadzu scanning electron microscope (SS-550, Kyoto, Japan), and high-definition SEM images of the samples were obtained.

To investigate the phase composition and crystallinity of the mineral components of each group of specimens, according to the X-ray diffraction (XRD) sampling requirements, multiple powder samples above 20 mg with a distance of approximately 25 mm from the surface of the specimen were collected. The samples were ground to within an order of magnitude of 40 μm, and a SmartLab SE instrument (Rigaku, Tokyo, Japan) was used to perform XRD. The scanning angle range of the instrument was set to 5~65°, the step size was set to 0.02° and the scanning speed was set to 4°/min to obtain the diffraction pattern of the samples. Then, JADE9.0 software was used for comparative analysis.

To understand the material composition of the test specimen and conduct quantitative analysis, according to the thermogravimetric (TG) analysis and sampling requirements, 10 mm × 10 mm × 10 mm specimens were collected after destruction, and a German NETZSCH STA-449C synchronous thermal analyzer (Thauern, Bavaria, Germany) was used to perform thermogravimetric treatment on the samples, which were preheated for 3 min after starting. The target temperature was set at 1000 °C, and the heating rate was 20 °C/min. The samples were placed in a ceramic crucible to obtain TG and DTG data.

3. Test Results and Analysis

3.1. Experimental Phenomenon

To investigate the 28-day cubic compressive strength values of the test specimens, all the specimens were tested on a yE-200A 2000 kN hydraulic pressure testing machine with displacement control and a loading rate of 0.02 mm/s. During the tests, cracks began to appear when the specimens were loaded to approximately 80% of their ultimate load, accompanied by a clear “ding-ding” sound from the peeling of the steel fibers. When a specimen was completely destroyed, multiple vertical cracks appeared, and the pressure curve dropped sharply. The fracture load threshold of the test specimen was 1100 kN, which was specifically expressed at the moment of complete failure. If the instantaneous load of the test specimen did not exceed 1100 kN, the specimen would have obvious cracks, and there were no obvious sounds. When the instantaneous load of the specimen exceeded 1100 kN, the failure of the test specimen was in a “semiburst” style, which was manifested as the specimen splitting into three or four pieces of gravel, accompanied by loud noise, but because of the existence of steel fibers, the gravel pieces were connected to each other and did not scatter around. The failure forms of each specimen in the compressive strength test at 28 days of age are shown in

Table 4. Failure forms of different specimens at the age of 28 d.

No.	U	G4	H3	GH4
Description of failure mode	There are many penetrating vertical cracks on the front and side.	There are two thick vertical through-cracks on the front, which divide the specimen into three pieces, and many crushed stones and steel fibers collapse around.	The test piece is broken into several pieces, and the whole piece has been seriously deformed. The internal volcanic stone is seriously broken, and there is a volcanic stone peeling phenomenon.	Similar to H3, but without H3, the overall deformation is serious, and the degree of peeling for the volcanic rock coarse aggregate is weaker than that for the H3 specimen.
Damage pattern picture

.

3.2. Test Results

In general, the cubic compressive strength of a specimen increased rapidly in the early stage and gradually decreased in the later stage, as shown in Figure 2. It is worth noting that the cubic compressive strength in the later stage for UHPC basically tended to be stable, but the cubic compressive strength in the later stage for UHPC with added MK still continuously improved.

At an age of 3 days, the cubic compressive strength values of the test specimens decreased in the order of G4 > G2 > G3 > U > G1. At an age of 7 days, the growth rate of the cubic compressive strength of G4 was significantly lower than that of the other groups, for which the cubic compressive strength decreased to the lowest level. The reason for this phenomenon is mainly the extremely low water–cement ratio of UHPC and relatively little free water. When cement and silica fume undergo a hydration reaction with free water at the same time, the MK activity of the high specific surface area is larger, and more free water is required for the hydration reaction, which affects its strength development. At 14 days of age, the growth rate of U‘s cubic compressive strength was significantly greater than that of the other groups, and its cubic compressive strength was highest, which indicates that there was more free water in U at this age, the cement hydration rate was the fastest and the growth rate of the cubic compressive strength was the highest. At 28 days of age, the growth rate of the cubic compressive strength of G4 was significantly greater than that of the other groups, the cubic compressive strength was the highest, reaching 136.78 MPa, and the cubic compressive strength of U was 123.14 MPa. The cubic compressive strength of G4 was 11.077% greater than that of U, and the cubic compressive strength of the specimens was in the order of G4 > G3 > U > G2 > G1, showing that the cubic compressive strength of the 28-day-old UHPC increased with an increasing MK dispersion ratio when the MK dispersion ratio was in the range of 4%.

By observing the cubic compressive strength curves of UHPC at ages of 3–28 days, it could be seen that the growth rate of the cubic compressive strength of G4 was the smallest during the entire process, and the cubic compressive strength curve was almost a straight line. The strength values at 3 and 28 days were the highest, which is obviously different from those of the other groups; the growth rates of the cubic compressive strength values in the early stage for all of the UHPC with MK added were lower than those of UHPC and the growth rate values of the cubic compressive strength in the later stage were higher than those of UHPC. This was because the MK particles absorbed part of the free water in the early stage of hydration, resulting in relatively slow hydration of the cement in the early stage of hydration, and the MK that absorbed free water in the later stage of hydration gradually underwent an active reaction, which will be further analyzed and explained in Section 2.2.

As shown in Figure 3a, at 28 days, the cubic compressive strength values of the test specimens were U > H3 > H2 > H1 in descending order, indicating that when the volcanic stone particle size was in the range of 9.5~31.5 mm, the cubic compressive strength of CA-UHPC increased with an increasing particle size. This was because the larger particle size of the volcanic stone aggregates led to a coarser network matrix composed of UHPC slurry, which caused the overall strength of the UHPC specimen to be higher. Moreover, volcanic stone has a higher crushing value (as much as 24.36% higher) than road basalt gravel. Therefore, although the compressive strength of CA-UHPC based on volcanic stone was lower than that of UHPC without coarse aggregates, the price of volcanic stone itself is very low, the compressive strength of volcanic stone CA-UHPC is still within the range of high-performance concrete (UPC) and it is still of high use value in the field of construction, while reducing the cost of building materials.

As shown in Figure 3b, at 28 days, the compressive strength of U was also the highest. The compressive strength of GH4 in CA-UHPC with MK was the highest, reaching 84.77 MPa, which is consistent with the test results for UHPC supplemented with MK, indicating that when the MK dispersion ratio was in the range of 4%, the cubic compressive strength of UHPC decreased first and then increased with an increasing MK dispersion ratio. The cubic compressive strength of CA-UHPC increased with an increasing MK dispersion ratio. The cubic compressive strength of GH4 was 34.27% greater than that of H1, and the improvement was more obvious than that of UHPC.

The stress curves for the CA-UHPC are similar; that is, there is a “weak segment” in the early stage, and then the stress curve gradually increases until failure. The reason for the “weak segment” is that there were occasional volcanic stone aggregates on the stress surface of a specimen. This part of the volcanic stone aggregate having first contact with the pressure testing machine resulted in a long pressure time history, and the curve thus produced had a “weak” effect.

4. Micro Analysis

4.1. SEM Image of the “Interlock Effect”

When the interface is magnified 200–300 times, Figure 4 shows that the interface between the volcanic stone and the concrete matrix was relatively dense, and the surface of the volcanic stone and the concrete matrix were well combined. Figure 5 clearly shows that the pores of the volcanic stone were filled with mortar, indicating that the pores of the volcanic stone better formed mechanical occlusive “interlocking” with the mortar matrix, which effectively enhanced the shear resistance performance of the volcanic stone–UHPC interface.

In addition, due to the porous structure of volcanic stone, the interface of the pore structure was the most unfavorable interface in the compressive process of the CA-UHPC, which was one of the main reasons for the decrease in the compressive strength of the CA-UHPC.

4.2. XRD Comparative Analysis

Figure 6 shows a comparison of the XRD patterns of the UHPC and CA-UHPC sample groups with added MK. Figure 6 shows that the reaction was mainly concentrated in the 25~35° section (2θ), forming an amorphous “wave crest” centered at 27~29° (2θ), in which the “wave crest” was the limestone phase. A small sharp peak was found at approximately 27°~30° (2θ), which is usually related to C-(A)-S-H gel or calcite formed by carbonatation. According to the standard diffraction pattern, it was determined that a C-S-H gel product was generated by the reaction.

Figure 6a–d shows that at different ages, the orders of the C-S-H diffraction peaks of each sample were basically the same as the orders of the compressive strength values. The diffraction peak of C-S-H represents the crystallinity of C-S-H, and the diffraction peak was proportional to the crystallinity, indicating that at the age of 28 days, the crystallinity of C-S-H was greater when the dispersion ratio of the MK was 4%, and the growth of the crystal plane was more orderly. A high dispersion ratio of MK promotes the crystallization of C-S-H, which can increase the strength of the material. The above analysis results verify the change rule of the cubic compression test described above.

Figure 6e shows that at 28 days, the C-S-H diffraction peak of the GH1 sample was significantly greater than those of the other dispersion ratios, indicating that the CA-UHPC with added MK at a dispersion ratio of 4% had a higher C-S-H crystallinity, which is consistent with the results shown in Figure 6d and also verifies the results of the previous cubic compression tests. In Figure 6, hydrargillite was found near 50.7° (2θ), and wustite was found near 60.7° (2θ). The intensities of the diffraction peaks at these two angles did not change significantly but decreased slightly in comparison in UHPC with added MK.

4.3. TG/DTG Analysis

Figure 7a–e are comparison diagrams of the TG and DTG curves, respectively, for the UHPC and CA-UHPC sample groups with added MK. The MK-added UHPC and CA-UHPC sample groups had obvious thermal decomposition reactions in the temperature ranges of 50~150 °C and 550~700 °C. The mass loss at 50~150 °C and 550~700 °C was caused by the evaporation of free water and bound water with an increasing temperature and the decomposition of some poorly crystallized CaCO₃ crystals to generate CO₂ with increasing temperature.

Comparing Figure 7a–d and Figure 7e, it can be seen that the mass loss rates of the CA-UHPC sample group with MK added were faster in the temperature ranges of 50~150 °C and 600~700 °C, which indicates that, on the one hand, the incorporation of volcanic stone coarse aggregate retains more free water and bound water, and on the other hand, it increases the crystallization of CaCO₃, and the decomposition temperature of CaCO₃ increased by 50 °C compared with that at other ages, indicating that the CaCO₃ crystals of the samples at 28 days of age were more stable as the crystal planes became more ordered. Figure 7e shows that the mass loss of GH4 in the temperature range of 600–700 °C was significantly greater than those of the other specimens, indicating that the amount of CaCO₃ crystals in GH4 was the greatest, and MK with a high dispersion ratio helped to generate more CaCO₃ crystals and was able to provide a higher strength. The results of the cubic compression tests described above also verified that MK with a high dispersion ratio had the most significant strength improvement for the CA-UHPC system.

Figure 7d shows that the decomposition temperature of ettringite is 70 °C, the weight loss at 70~150 °C was basically the weight loss of the bound water in the C-S-H gel and the ettringite and the weight loss ratios were 2.97%, 2.99%, 3.16% and 3.71%, respectively. At 28 days of age, the masses of C-S-H gel and ettringite in G4 were approximately 1.24 times greater than those in G1, and the cubic compressive strength of G4 was approximately 1.22 times greater than that of G1, which are basically equal ratios. The TGA results further verified the results of the cubic compressive strength test described above.

The decomposition temperatures of Ca(OH)₂ and CaCO₃ crystals with poor crystallization and CaCO₃ crystals with better crystallization are 430–550 °C, 550–750 °C and 750–950 °C, respectively. Figure 8 shows that there was a relatively obvious “concave section” in the temperature ranges of 50–150 °C and 600–700 °C in the UHPC and CA-UHPC sample groups with MK added, which is consistent with Figure 7, and there was a smaller “concave section” at approximately 430 °C. After analysis, it was found that the “concave section” in the temperature range of 50~150 °C was due to the acceleration of the weight loss caused by the evaporation and the loss of bound water in the C-S-H gel (CaO·2SiO₂·3H₂O) and ettringite (3CaO·Al₂O₃·3CaSiO₄·32H₂O). The “concave section” in the temperature range of 600~700 °C was due to the acceleration of weight loss caused by the decomposition of some poorly crystallized CaCO₃ crystals with an increasing temperature to produce CO₂, which is consistent with the analysis of the TG curve. The “concave section” near 430 °C was caused by the conversion of Ca(OH)₂ into CaO and free water with increasing temperature and the loss of free water with high-temperature evaporation.

As shown in Figure 8a–d, it can be observed that G4 was at the lowest point in the “concave section” in the temperature range of 50~150 °C, indicating that the contents of C-S-H and ettringite in the G4 specimen at the age of 28 days were relatively greater than those in the other UHPC specimens with added MK. Figure 8e shows that the position of GH4 was significantly lower those that of the other specimens in the concave section in the temperature range of 600–700 °C, indicating that the CaCO₃ content in the GH4 specimen at the age of 28 days was significantly greater than those in the other CA-UHPC specimens with added MK and different dispersion ratios. High contents of C-S-H, ettringite and CaCO₃ led to the higher compressive strength values for the G4 and GH4 specimens, which was verified by the results of the cubic compressive test described above.

Comparing the TG curves of the MK-added UHPC sample groups at different ages, it can be seen that the “concave section” of the UHPC samples with a 4% MK dispersion ratio added at 3 days and 28 days of age was the most distinctive in the temperature range of 600–700 °C, indicating that the CaCO₃ contents of UHPC with a 4% MK dispersion ratio added were greater at 3 and 28 days of age, which is consistent with the results on the cubic compressive strength.

5. Conclusions

(1)

Volcanic rocks as a new type of coarse aggregate for UHPC were used not only to maintain the compressive strength of UHPC but also to significantly reduce the construction cost of UHPC projects, resulting in the development of a new type of high-strength engineering material. The conclusions of this research provide reliable data support for the UHPC theoretical framework.

(2)

The volcanic stone pore structure and the mortar matrix form a mechanical “interlocking effect” structure, which can effectively improve the interfacial shear resistance of CA-UHPC. At the same time, the filling interface of volcanic stone pores was one of the main reasons for the decrease in the compressive strength of CA-UHPC, and the combined effect of the two was that the latter dominated.

(3)

The crystallinity of C-S-H was greater for both UHPC and CA-UHPC with MK added at a dispersion ratio of 4%. MK at a content of 4% promoted the crystallization of C-S-H, which increased the strength of UHPC. On the one hand, the incorporation of volcanic stone coarse aggregate retained more free water and bound water, and on the other hand, it also increased the crystallization of CaCO₃. The combined effect of MK with a high dispersion ratio and volcanic stone promoted the crystallization of CaCO₃.

(4)

When the MK dispersion ratio was less than 4%, the contents of CaCO₃ crystals, C-S-H gel and calcium vanadium increased with an increasing MK dispersion ratio. At the age of 28 days, an MK dispersion ratio of 4% improved the quality of the C-S-H gel and ettringite by 24.82% compared with an MK dispersion ratio of 1%. A high content of C-S-H, ettringite and CaCO₃ caused the cubic compressive strength of the GH4 specimen to be higher.