Influence of Random Vehicle–Bridge Coupling Vibration on the Anti-Disturbance Performance of Concrete Materials

Abstract

Without interrupting the traffic on the old bridge, the connection of the widening bridge will cause disturbance to the concrete in the splicing position. In order to study the anti-disturbance performance of concrete material, the Normal Concrete (NC) material and the Ultra-High-Performance Concrete (UHPC) material were experimentally investigated by the vertical shaking table and X-ray computed tomography scanning. It can be learnt from tests that the compressive strength of NC and UHPC can be increased by about 10% to 20% after disturbance, while the flexural tensile strength and splitting tensile strength of both NC and UHPC can be reduced by about 20% to 25% and 10% to 20%, respectively. The elastic modulus of UHPC is not significantly affected by the vibration disturbance, and that of NC can be increased by about 20%. The setting time difference of the proposed NC material can be controlled within 100 min, and it can improve its anti-disturbance performance. Excessive vibration disturbance affects the internal structure of NC, while it has little effect on the distribution of steel fiber in UHPC. Due to the high cost of UHPC materials, it is recommended to analyze the joint performance requirements before the selection of splicing materials. If the stress requirement is not particularly high, it is still recommended to apply the proposed NC material for the splicing of widening bridge.

1. Introduction

In bridge reconstruction and extension projects, the following three splicing forms are generally adopted during the connection of the old existing bridge and the new widening bridge [1,2]: superstructure connection and substructure separation (C-S), both superstructure and substructure separation (S-S) and superstructure connection and substructure connection (C-C), as shown in Figure 1. The first splicing form (C-S) is usually used in the actual project to avoid the differential settlement between the old and new part of the widening bridge [3,4]. However, the traffic on the old existing bridge is always required to be uninterrupted during the process of superstructure splicing. Due to the traffic still operating on the old existing bridge, the phenomenon of vehicle–bridge coupling vibration will occur on the superstructure, while the new widening bridge is relatively static [5]. When pouring concrete in the splicing position, vibration disturbance will act on the concrete, and the material performance will be affected [6,7,8].

In order to improve the performance of concrete at the splicing position, several studies have been carried out on the control of vehicle–bridge coupling vibration. Corresponding measures are put forward from three aspects of structural deformation control, construction method optimization and concrete performance improvement.

In regard to the aspect of controlling the structural deformation, Kwan et al. recommended several possible methods to reduce damage on concrete stitches in the bridge-widening project, such as temporary shear connections, temporary propping, segmental concrete stitching and so on [9]. Although these methods play a great role in reducing bridge vibration, the concrete tends to crack after the removal of temporary measures. Chen proposed a new method of controlling the deflection difference between the existing and new bridge based on the tuned mass damper (TMD) [10]. However, the principle of this equipment is too complex to popularize in practical engineering projects.

Guo et al. also explored this issue from the aspect of optimizing construction methods as well. Three construction schemes were examined and compared with vibration tests and static tests [11]. Casting twice with steel beams was recommended for practical engineering applications. However, the widening splicing through improving construction measures will still have complex process problems and affect the construction progress.

Improving the performance of concrete is considered the most convenient and economical measure [12,13]. However, before developing new materials or optimizing existing materials, it is still necessary to clarify the performance degradation of existing materials under vehicle–bridge coupling vibration, and to determine the optimization direction of widening joint materials [14]. Fossetti et al. conducted an experimental investigation on concrete subjected to induced vibrations, and the test results showed that compression strength can be degenerated by vibrations while tensile behavior will not be apparently affected [15].

In this article, the anti-disturbance performance of concrete materials is experimentally investigated based on an actual bridge extension project. The finite element model was established based on the theory of random vehicle–bridge coupling vibration, and the test parameters of the disturbed concrete at the splicing position are determined. The vibration disturbance at the splicing position was simulated by the vertical shaking table. The anti-disturbance property of Normal Concrete (NC) and Ultra-High-Performance Concrete (UHPC) are compared and analyzed.

The finite element model is established according to the actual project, and the numerical calculation is carried out based on the random vehicle–bridge coupling vibration theory. It can make the vibration test parameters adopted in this paper more realistic. The vertical shaking table is applied in this test, and the long-term intermittent vibration mode is adopted as well. Based on the determined test parameters, the simulation of vibration disturbance at the splicing position under the circumstance of no interruption of traffic can be realized. By using different types of standard specimen, the anti-disturbance performance of NC and UHPC are comprehensively compared and analyzed, which provides ideas for the development of anti-disturbance concrete materials at the splicing position.

2. Vibration Characteristic Analysis of Splicing Region

2.1. Damage Source of Disturbed Concrete in Expansion Bridge Splicing Region

In the bridge reconstruction and extension project, the old existing bridge and the new widening bridge are relatively independent before the widening connection. Dynamic deflection deformation will be caused by vehicle vibration on the old existing bridge, while the new widening bridge is still in a static state. Therefore, differential vertical deformation will appear at the position of the splicing region. In the process of concrete curing, aggregates of concrete will be disturbed by vibration as shown in Figure 2, and concrete performance degradation will occur.

In such cases, the damage degree of concrete is related to the vibration characteristics of the bridge to a certain extent. Therefore, it is necessary to clarify the bridge vibration response of the concrete in the splicing joint.

2.2. Simulation Method Based on Random Vehicle–Bridge Coupling Vibration

The classical half-vehicle model is adopted to describe the vibration characteristics of the vehicle. Based on the modal synthesis method and the mode superposition method, the modal synthesis equation of the bridge structure with generalized coordinates and vehicle degrees of freedom is established as shown in Equation (1).

$$\begin{array}{l}\left(\begin{array}{cc}\left[I\right]& \left[0\right]\\ \left[0\right]& \left[M_v\right]\end{array}\right)\left(\begin{array}{c}\left\{{\ddot{q}}_r\right\}\\ {\ddot{Z}}_v\end{array}\right)+\left(\begin{array}{cc}\left[2\omega \xi \right]+{\left[\mathsf{\Phi }\right]}^T{\left[N\right]}^T{c}_{ti}\left[N\right]\left[\mathsf{\Phi }\right]& -{\left[\mathsf{\Phi }\right]}^T\left[N\right]c_{ti}\\ -\left[\mathsf{\Phi }\right]\left[N\right]c_{ti}& \left[C_v\right]\end{array}\right)\left(\begin{array}{c}\left\{{\dot{q}}_r\right\}\\ {\dot{Z}}_v\end{array}\right)\\ +\left(\begin{array}{cc}\left[{\omega }^2\right]+{\left[\mathsf{\Phi }\right]}^T{\left[N\right]}^T{k}_{ti}\left[N\right]\left[\mathsf{\Phi }\right]& -{\left[\mathsf{\Phi }\right]}^T\left[N\right]k_{ti}-V{\left[\mathsf{\Phi }\right]}^T\left[N_x\right]c_{ti}\\ -\left[\mathsf{\Phi }\right]\left[N\right]k_{ti}& \left[K_v\right]\end{array}\right)\left(\begin{array}{c}\left\{q_r\right\}\\ Z_v\end{array}\right)=\left(\begin{array}{c}{\left[\mathsf{\Phi }\right]}^T\left\{F_{bv}^{VIB}\right\}\\ F_v^G+F_{bv}^{VIB}\end{array}\right)\end{array}$$

(1)

in which $\left[I\right]$ denotes the n-order unit matrix; $\left[M_v\right]$, $\left[{\mathrm{C}}_v\right]$ and $\left[K_v\right]$ are the mass of the vehicle and the damping and stiffness matrix of the vehicle suspension, respectively; $\omega$ and $\xi$ are a certain order frequency and damping ratio of the bridge structure, respectively; $\left[\mathsf{\Phi }\right]$ is the modal coordinate corresponding to the frequency of the bridge structure; $\left[N\right]$ is the interpolation function for the vehicle tire movement to the bridge structure; $c_{ti}$ and $k_{ti}$ represent the vehicle tire damping and stiffness, respectively; $F_{bv}^{VIB}$ is the interaction force matrix between the vehicle and the bridge; $F_v^G$ is the force matrix assigned to the wheel of the vehicle; $\left\{{\ddot{q}}_r\right\}$, $\left\{{\dot{q}}_r\right\}$ and $\left\{q_r\right\}$ are the acceleration, velocity and displacement vectors under the r-order modal coordinates of the bridge structure, respectively; ${\ddot{Z}}_v$, ${\dot{Z}}_v$ and $Z_v$ are response matrices of the vehicle acceleration, velocity and displacement, respectively.

In addition, the pavement roughness is considered as well, and the function adopted is shown as Equation (2).

$$r(x)={\displaystyle \sum _{k=1}^N\sqrt{2\Delta n\phi \left(n_0\right){\left(\frac{n_k}{n_0}\right)}^{-W}}\cos \left(2\pi n_{k}x+{\theta }_k\right)}$$

(2)

where $x$ is the horizontal displacement of pavement, $\phi \left(n_0\right)$ is the pavement roughness coefficient, $n_0$ is the reference frequency, $W$ is the frequency index and ${\theta }_k$ is a random number.

Cellular Automata (CA) analysis theory is introduced to simulate the random traffic flow, and the analysis model is established with MATLAB [16]. The traffic lanes of the bridge are discretized into continuous cellular spaces, and the function for describing pavement roughness is integrated into the cell space [17]. The process of random vehicle–bridge coupling vibration simulation is shown in Figure 3. The modal matrix and frequency of bridge structure can be obtained by finite element analysis. The modal vector is interpolated and stored into the cell space together with the structural frequency. The known vehicle characteristics and traffic characteristics parameters are used to update the vehicle position in real time in the analysis model. Through the change of vehicle position, the simulation of random movement of vehicle axles on bridges can be realized. Finally, the randomly generated vehicle queue is arranged in the cell space. The simulation of vehicle–bridge coupling vibration can be realized along with time.

2.3. Vibration Response Analysis of the Actual Bridge

The Second Danshui River bridge in the West Shen-shan expressway is selected. It is a T-beam widening bridge with span length of 25 m as shown in Figure 4. At the splicing region, the flange plates and the diaphragm plate are connected by concrete.

The Second Danshui River bridge is selected as an actual bridge case to obtain its modal matrix and structural frequency. In the bridge reconstruction and extension project, the old existing bridge and the new widening bridge are relatively independent before the widening connection. It can be considered that the dynamic deflection deformation caused by vehicle vibration on the old existing bridge cannot affect the new widening bridge. That is to say, the new widening bridge has no obvious dynamic deformation before the widening connection.

Therefore, only the finite element model of the existing old bridge was established, and it was applied to analysis the relational deformation difference at the splicing position. The established model is shown as Figure 5. It can be learnt from the analysis result of the finite element model that the first order frequency of the bridge is 4.31927 rad/s.

Based on the statistics of natural frequency and model, vibration response analysis of the bridge is carried out based on the theory of random vehicle–bridge coupling vibration. In this model, the surface roughness level is considered as Level B, and only one fast lane and one slow lane are opened on the bridge. The random vehicle–bridge coupling vibration simulation results is presented in Figure 6.

According to the bridge vibration response simulation result, the bridge structure shows obvious flexural deflection under the vibration action, and the structural total deformation is a range of 3 mm. The time period 8:00 to 20:00 is regarded as the daytime period, and 20:00 to 0:00 and 0:00 to 8:00 are regarded as the night-time period. In the daytime period, the structural deformation is mainly concentrated around −4 mm. The vibration deflection caused by traffic at night is slightly smaller than that at daytime, and it is mainly about −1 mm.

3. Test Setup

3.1. Concrete Mix Proportion

In order to study the influence of random vehicle–bridge coupling vibration on concrete at the splicing position, and to evaluate the maintenance degree of concrete material after disturbance, experimental studies were carried out on concrete materials with different mix proportions. Normal Concrete (NC) and Ultra-High-Performance Concrete (UHPC) were experimental investigated in this study.

• Normal Concrete (NC)

Two types of normal concrete including NC-1 and NC-2 were mainly tested in this study. Normal concrete NC-1 is mainly composed of cement, stone, sand, water and superplasticizer. The mix proportion of NC-1 is shown in

Table 1. Mix proportion of NC.

Type		NC-1	NC-2
Mix proportion (kg/m3)	Cement	485	400
	Fly ash	/	100
	Stone	1173	990
	Sand	632	810
	Water	150	165
	Superplasticizer	4.85	5
Sand ratio		0.35	0.45
Actual unit weight (kg/m3)		2350	2300

. Relevant studies have shown that fly ash can improve the fluidity, cohesiveness and water retention of concrete mixture, and that it can make concrete mixture easy to be pumped and poured as well [7,18]. Therefore, fly ash is added to the normal concrete on the basis of mix proportion of NC-1, and the amount of coarse aggregate is also optimized. Therefore, another type of normal concrete NC-2 is proposed, and the mix proportion of NC-2 is shown in Table 1 as well.

Portland cement P·II 52.5 was adopted in normal concrete. The specific surface of cement is 374 m²/kg and its loss on ignition is 2.41%. The initial and final setting time of cement were 155 min and 210 min, respectively. The grade of fly ash is classified as level II according to the Chinese standard GB/T 1596 [19]. The chemical composition of cement and fly ash were analysed according to the Chinese standard GB/T 176 [20], and the test results are shown in

Table 2. Chemical analysis result of cement and fly ash (wt./%).

Chemical Composition	Cement	Fly Ash
CaO	78.91	4.59
SiO2	10.66	4523
Fe2O3	4.42	4.63
SO3	3.3	0.81
Al2O3	1.24	39.56
K2O	0.75	1.03
TiO2	0.3	1.58
MnO	0.19	0.06
MgO	0.08	0.97
Loss	0.15	1.54

. The adopted sand is natural river sand of which the technical requirement can meet Grade II. In addition, its fineness module is 2.4, which can be classified as medium sand. Sika^® high performance polycarboxylate superplasticizer is adopted, and its solid content is 20%.

• Ultra-High-Performance Concrete (UHPC)

The investigated Ultra-High-Performance Concrete is also including two types, UHPC-1 and UHPC-2, and the mix proportion of UHPC is shown in

Table 3. Mix proportion of UHPC.

Type	UHPC-1	UHPC-2
Powder (kg)	25	25
Steel fiber (kg)	2.4	2.2
Water (kg)	2.425	2.32

.

Among them, UHPC-1 is in the form of separate packaging of powder and steel fiber, which needs to be mixed with water after mixing powder and steel fiber. Conversely, UHPC-2 is a pre-mixed material of powder and steel fiber, which can be directly used after mixing water. The diameter of steel fiber in UHPC-1 is 0.2 mm and the average length is 13 mm. The diameter of steel fiber in UHPC-2 is 0.2 mm and the average length is 17 mm.

3.2. Specimen Classification

Differential vibration deformations exist in the widened bridge during its splicing process, and the concrete in the widening joint will be affected by disturbance before setting. The test conditions with different vibration amplitudes and frequencies were experimentally investigated to study the anti-disturbance performance of different concrete materials after the vibration.

According to the vibration characteristic simulation analysis result of the Second Danshui River bridge and corresponding research results [15], the vibration characteristics simulated by the finite element model are as follows:

• Vibration amplitude range: 0–4.0 mm, and the mean value is about 3 mm.
• Vibration frequency range: 0–9 Hz, and the mean value is about 4 Hz. In addition, the possibility of particularly high frequency vibration in the test is low as well.

Therefore, Materials NC-1 and UHPC-1 were tested with a vibration amplitude of 0–4 mm and a vibration frequency of 3–9 Hz, respectively. Materials NC-2 and UHPC-2 were only tested under the vibration parameters of 3 mm and 6 Hz, and the test condition of 0 mm and 0 Hz was also adopted for contrast. In this test, the classification and test parameters of each group of specimens are shown in

Table 4. Specimen classification and test parameters.

Serial Number of Specimen Group	Material	Vibration Parameter		Disturbance Form	Specimen Type
		Amplitude (mm)	Frequency (Hz)
NC1-0-0	NC-1	0	0	Static state	Contrast specimen
NC1-2-6		2	6	Vertical vibration	Disturbed specimen
NC1-3-3		3	3
NC1-3-6		3	6
NC1-3-9		3	9
NC1-4-6		4	6
NC2-0-0	NC-2	0	0	Static state	Contrast specimen
NC2-3-6		3	6	Vertical vibration	Disturbed specimen
UHPC1-0-0	UHPC-1	0	0	Static state	Contrast specimen
UHPC1-2-6		2	6	Vertical vibration	Disturbed specimen
UHPC1-3-3		3	3
UHPC1-3-6		3	6
UHPC1-3-9		3	9
UHPC1-4-6		4	6
UHPC2-0-0	UHPC-2	0	0	Static state	Contrast specimen
UHPC2-3-6		3	6	Vertical vibration	Disturbed specimen

.

3.3. Test Device

A vertical shaking table was employed to simulate the vertical vibration disturbance on the specimens. The model of vertical shaking table adopted is ZD/YH-F 600 from Yihua^® Instrument Co., Ltd., Shanghai, China. The mesa size of the shaking table is 60 cm × 60 cm. The maximum load capacity of the vertical shaking table is 400 kg, the vibration amplitude range is 0–10 mm, and the vibration frequency range is 0–999 Hz.

In addition, a 120 s× 120 cm wooden test platform was fabricated and fixed on the vertical shaking table, as shown in Figure 7. The specimens falling from the vertical shaking table during the test can be avoided.

Standard specimens mold for different test purposes were adopted according to the Chinese standard GB/T 50081 and GB/T 50082 [21,22]. Each group of NC specimens and UHPC specimens contain different types of standard specimen as listed in

Table 5. Different standard specimen molds in different specimen groups.

No.	Material	Purpose	Quantity	Mold Internal Size (mm)	Shape
1	NC *	Compressive strength	3	150 × 150 × 150	cube
2		Flexural tensile strength	3	150 × 150 × 600	cuboid
3		Splitting tensile strength	3	150 × 150 × 150	cube
4		Elastic modulus	6	150 × 150 × 300	cuboid
5		Shrinkage of concrete	3	100 × 100 × 515	cuboid
1	UHPC *	Compressive strength	3	150 × 150 × 150	cuboid
2		Flexural tensile strength	3	150 × 150 × 600	cuboid
3		Axial tensile strength	4	φ100 × 100	dumbbell shaped
4		Elastic modulus	6	150 × 150 × 300	cuboid
5		Shrinkage of concrete	3	100 × 100 × 515	cuboid

* NC: Normal Concrete; UHHP: Ultra-High-Performance Concrete.

.

3.4. Experimental Process

The preparation of each group of specimens was mainly conducted according to the following process.

Firstly, normal concrete and ultra-high-performance concrete were respectively mixed and poured into the corresponding mold before its initial setting. Test parameters of the vertical shaking table were set according to different specimen groups.

Secondly, the same series of specimen were fixed in the wooden test platform. After the fixation was completed, the specimen was subjected to vibration disturbance according to the proposed vibration parameters. Taking 1 min as a disturbance period, the specimen was vibrated for 15 s and then pause for 45 s in a period. The cumulative vibration time lasted until 18 h after the concrete was cast.

Thirdly, the specimens were cured in static state. Under the circumstance of (20 ± 5) °C and RH ≥ 50%, the specimen was maintained within the mold for 30 h. After the mold was removed, it was maintained in the environment of (20 ± 2) °C and RH ≥ 95% for 26 days. That is to say, the cumulative curing time of the concrete after pouring is 28 days.

Finally, specimens were loaded by the universal testing machine according to the Chinese standard GB/T 50081 and GB/T 50082 [21,22]; different strength parameters and elastic modulus can be obtained. In addition, the concrete shrinkage parameter was tested by the non-contact concrete-shrinkage and -deformation measuring instrument.

4. Influence of Vehicle–Bridge Coupling Vibration on Anti-Disturbance Performance of Concrete

4.1. Influence on Aggregate Segregation

The aggregate segregation rate of each normal concrete material after vibration disturbance is tested. The test results are shown in

Table 6. Aggregate segregation rate of NC specimens.

Serial Number of Specimen Group	Vibration Parameter		Rate of Aggregate Segregation (%)
	Amplitude (mm)	Frequency (Hz)
NC1-0-0	0	0	1.6
NC1-2-6	2	6	1.7
NC1-3-3	3	3	1.7
NC1-3-6	3	6	1.8
NC1-3-9	3	9	1.7
NC1-4-6	4	6	2.0
NC2-0-0	0	0	2.2
NC2-3-6	3	6	3.2

.

It can be found from the test results that there is no significant difference in aggregate segregation rate of normal concrete under vibration disturbance conditions. The upper and lower concretes show good uniformity; no obvious bleeding or segregation phenomena were found on the surface of concrete during the test. In general, the proposed NC material can significantly shorten the setting time, which is beneficial to reduce the segregation degree of coarse aggregate.

4.2. Influence on Concrete Strength

The normal concrete (NC-1 and NC-2) and the ultra-high-performance concrete (UHPC-1 and UHPC-2) were vertically disturbed under the different random vehicle–bridge coupling simulation vibration circumstances. The physical and mechanical properties of the adopted concrete material were tested according to Chinese standard GB/T 50,081 [21]. Different types of test result are compared as shown in

Table 7. Mechanical properties of different specimen groups (28 d *, Unit: MPa).

Serial Number	Compressive Strength	Flexural Tensile Strength	Splitting Tensile Strength	Elastic Modulus Under Static Compressive Stress
NC1-0-0	52.7	6.1	5.17	58.64
NC1-2-6	57.7	6.0	5.30	47.43
NC1-3-3	59.9	4.8	4.71	49.57
NC1-3-6	61.7	5.9	4.08	51.11
NC1-3-9	61.1	4.9	4.74	46.16
NC1-4-6	62.0	4.6	4.89	47.39
NC2-0-0	54.1	5.1	4.30	43.40
NC2-3-6	52.6	5.8	3.98	47.50
UHPC1-0-0	111.3	28.0	10.21	46.90
UHPC1-2-6	115.6	24.0	8.98	44.71
UHPC1-3-3	134.7	24.6	8.89	43.93
UHPC1-3-6	125.1	22.9	8.16	45.65
UHPC1-3-9	130.6	17.0	6.91	46.28
UHPC1-4-6	137.5	20.1	8.20	46.88
UHPC2-0-0	124.1	27.1	6.80	46.23
UHPC2-3-6	118.4	21.0	6.59	46.93

* 28 d: Test value of specimen after 28 days of curing.

.

Based on the test results of the static specimens (NC1-0-0, NC2-0-0, UHPC1-0-0 and UHPC2-0-0), the test results of other specimens are normalized, as shown in Figure 8.

• Compressive strength

From the test results of the NC specimen, it can be learnt that the vehicle–bridge coupling vibration has a positive effect on the compressive strength of NC-1 and NC-2 in the range of test vibration amplitude and frequency. The compressive strength can be enhanced by 10% to 20%, which indicated that vibration disturbance can improve the density of normal concrete and is beneficial to the improvement of compressive strength of concrete.

From the test results of UHPC specimen, it can be learnt that the compressive strength of UHPC can be improved under the test vibration amplitude and frequency. It is mainly due to that the vibration can make UHPC material denser, which is conducive to the improvement of compressive strength.

• Flexural tensile strength, splitting tensile strength and elastic modulus

For the NC specimens, a negative effect occurs on the flexural tensile strength, splitting tensile strength and elastic modulus under static compressive stress after the vibration disturbance. It is mainly because these parameters are sensitive to the characteristics of the interface transition zone between coarse aggregate and cement paste. Existing studies have indicated that the vibration from the initial setting to final setting will lead to mechanical property deterioration of this interface transition zone [23,24]. Macroscopically, it is reflected in the decrease in flexural tensile strength, splitting tensile strength and the elastic modulus under static compressive stress. The test results show that the flexural tensile strength and static compressive elastic modulus decrease about 20% under the disturbance conditions of 4 mm–6 Hz and 3 mm–9 Hz.

However, for the UHPC specimens, the elastic modulus under static compressive stress is not significantly affected, while the flexural tensile strength and axial tensile strength are degenerated greatly. The main reason for this phenomenon is that the vibration disturbance may change the distribution direction of steel fiber and reduce the adhesive force between the steel fiber and the matrix of UHPC as well. It can also be learnt from the test results that high vibration frequency has an adverse influence on the mechanical property of UHPC. With the increase in vibration frequency, the loss of flexural tensile strength and axial tensile strength will be significantly accelerated. This may be due to the adverse effect of the increasing frequency on the distribution of steel fiber and its adhesive strength with the UHPC paste [25].

4.3. Influence on Slump and Setting Time

The tested slump of NC-1 is 200 mm, and the initial setting time and final setting time are respectively 305 min and 400 min. In addition, the retention rate of 28 d flexural tensile strength of NC-1 is 97% under the vibration condition of 3 mm and 6 Hz. The initial and final setting time of concrete reported in the relevant research are 755 min and 960 min, respectively. In addition, its retention rate of 7 d flexural tensile strength was 86.7% under the vibration amplitude of 3.5 mm and frequency of 5 Hz. By comparing with the normal concrete NC-1 in this article, it can be concluded that the loss of flexural tensile strength can be effectively reduced by shortening setting time.

Therefore, the fly ash is added into the concrete NC-1 and another type of normal concrete NC-2 is developed to reduce the setting time. Similarly, the normal concrete NC-2 were also tested under the static states and vertical vibration disturbance status. The slump of NC-2 is 200 mm, and the initial setting time and final setting time are 295 min and 390 min, respectively. The setting time of NC-2 is reduced to some extent by comparing with the NC-1. The property mechanical test results of NC-2 are also shown in Table 7. It can be learnt from the test results that flexural tensile strength and elastic modulus under static compressive stress of the two NC material are increased by around 10%, and the loss of splitting tensile strength is relatively small. By comparing the test result of NC material, it can be concluded that increasing the sand ratio and the amount of stone can effectively improve the anti-disturbance performance of concrete.

The tested slump of UHPC is about 500 mm to 700 mm, and it has better fluidity compared with the Normal Concrete material. In addition, the initial and final setting time of UHPC are 300 min and 840 min, respectively. Its initial setting time is not much different from that of the proposed NC materials, which can also reduce the effect of disturbance on the properties of concrete materials by a shorter setting time.

4.4. Influence on Concrete Shrinkage

Tests for shrinkage of NC and UHPC material were conducted under the static state and vibration disturbance condition according to Chinese standard GB/T 50082 [22], and the test result is shown in Figure 9.

It can be learnt from Figure 9a that the vehicle–bridge coupling vibration has little effect on the shrinkage property of NC material under the vibration disturbance condition of 3 mm and 6 Hz. Additionally, after optimizing the amount of coarse aggregate, the shrinkage ratio of the NC-2 specimen (NC2-0-0) increases under the static state, and the shrinkage of NC-2 specimen (NC2-3-6) under vibration disturbance condition decreases significantly. It can be concluded that the shrinkage property of NC material is sensitive to the amount optimization of coarse aggregate.

It can be learnt from Figure 9b that the vehicle–bridge coupling vibration is not significantly affected by the shrinkage performance of UHPC material under the vibration condition of 3 mm and 6 Hz. In particular, the shrinkage ratio of UHPC-1 is larger than that of UHPC-2 under both static and vibration disturbance conditions. For UHPC-1 material, the shrinkage value of specimens poured under vibration disturbance condition is less than that of specimens poured under static state (UHPC1-0-0) within 45 days curing time. However, it exceeds the specimens poured under static condition (UHPC1-0-0) after the 60 days, while for the shrinkage value of UHPC-2 material, specimens poured under vibration disturbance condition is always less than that of specimens poured under static state.

4.5. Influence on Concrete Compactness

Ultrasonic test was conducted on each group of NC specimens according to Chinese standard T/CECS 02 [26], and the test results are shown in Figure 10.

It can be learnt from Figure 10a that the sound velocity of concrete specimen can be improved with the coupling vibration effect of vehicles and bridge. In addition, it can be considered that vibration can improve the density of concrete. However, when the vibration amplitude increases to 4 mm or the frequency increases to 9 Hz, the ultrasonic velocity test result of NC specimens decreased. It indicated that excessive amplitude or frequency is unfavourable to the internal structure of concrete.

While for the ultrasonic test results of NC-2 specimens shown in Figure 10a, it can be seen that the ultrasonic velocity of the upper part and the lower part of the concrete sample under either static state or vibration disturbance conditions (3 mm and 6 Hz) are roughly the same. It can be known that the density of NC material can be increased after optimizing the amount of coarse aggregate. In addition, good uniformity of NC-2 material can be presented, and the anti-disturbance performance of NC-2 material can be better as well.

It can be seen from Figure 10b that the vehicle–bridge coupling vibration has little effect on the ultrasonic sound velocity of UHPC under different vibration frequency and amplitude. It is mainly due to that the two kinds of UHPC adopted in this test are Materials with low porosity and high density. Vibration disturbance within the range of vibration parameters adopted in this test does not significantly change the compactness of UHPC specimen. While for the lower part of the specimen UHPC2-3-6, the slight decrease in ultrasonic sound velocity is mainly cause by the uneven distribution of steel fiber.

4.6. Influence on Fiber Distribution

The steel fiber distribution of UHPC specimens is presented in Figure 11. It can be observed from Figure 11 that vibration makes steel fibers deposit at the bottom of the UHPC specimen. Vibration does have a great effect on the steel fiber distribution of UHPC. At the same time, steel fibers tend to be oriented, and the proportion of steel fibers distributed vertically increases. It will reduce the flexural tensile strength and axial tensile strength of UHPC.

In addition, X-ray Computed Tomography (XCT) scanning has also been conducted on the fiber distribution of each specimen. The space position coordinates of steel fiber can be determined through the XCT scanning, the angle of steel fiber with Z-axis (perpendicular to the direction of forming surface) can be calculated and the angle distribution of steel fiber can be analysed. The XCT test results of UHPC1 and UHPC2 are analysed as respectively shown in Figure 12, Figure 13 and Figure 14. It can be learnt from Figure 12 that the proportion of steel fibers perpendicular to the Z-axis decreased. This indicates that vibration can change the orientation of steel fiber. It will reduce the flexural strength and axial tensile strength, which is consistent with the results of macroscopic mechanical tests.

Similarly, UHPC2 also has the phenomenon of steel fiber orientation which is more obvious. At the same time, serious delamination and shrinkage of steel fiber occurred, and the number of steel fibers in the upper layer decreased (Figure 13 and Figure 14). This conforms to the phenomenon that the flexural strength and axial tensile strength of UHPC2 are significantly lower than those of UHPC1 in macroscopic experiments.

4.7. Synthetically Comparison of Concrete Material Properties

By comparing the properties of NC and UHPC materials under the influence of different vibration parameters, it was found that vibration disturbance has a positive effect on the compressive strength, but it has negative effect on the tensile strength. Shortening the setting time of NC can reduce the loss of tensile strength, and the anti-disturbance performance of NC can be effectively improved by enhancing the sand ratio and the amount of stone.

Due to the high tensile strength of UHPC material, it can still maintain a higher tensile strength than that of NC material after disturbance. It has a high performance advantage when applied to the splicing region of expansion bridge project. However, the UHPC material adopted is expensive, which is about ten times that of the NC material proposed. It is necessary to analyze the mechanical performance of the joint before selecting the material. After analysis and calculation, if the force requirement is not particularly high, it is still recommended to use the proposed NC material as the connection material for the expansion bridge splicing region.

5. Conclusions

The influence of different vibration patterns on the anti-disturbance performance of NC and UHPC material were compared. The main findings can be concluded as follows:

• Vibration disturbance has a positive effect on the compressive strength of both NC and UHPC, and the compressive strength can be increased by about 10% to 20% after disturbance. While vibration disturbance has a negative effect on the flexural tensile strength and splitting tensile strength of both NC and UHPC, and will be reduced by about 20% to 25% and 10% to 20%, respectively. In addition, the elastic modulus of UHPC is not significantly affected by the vibration disturbance, but that of NC can be increased by about 20%. Overall, the performance of UHPC is still much higher than that of NC, but the material cost of UHPC is relatively high.
• The loss of flexural tensile strength of NC can be effectively reduced by shortening the setting time. The time difference between the initial setting and the final setting of the proposed NC material can be controlled within 100 min, which is extremely beneficial to improving the anti-disturbance performance of NC material. The shrinkage property of NC is sensitive to the amount of coarse aggregate, and the anti-disturbance performance of NC can be effectively improved by enhancing the sand ratio and the amount of stone.
• Excessive amplitude or frequency is unfavorable to the internal structure of NC. The compactness and the shrinkage property of UHPC cannot be significantly changed by the vibration disturbance within the range of vibration parameters adopted in this test.