Local impact factor assessment for deck slabs of steel–concrete composite bridge system with twin girders

Abstract

The steel–concrete composite bridge system with twin girders, referred to as a steel plate composite girder bridge, is widely adopted for short- to medium-span highway bridges due to its ability to enable rapid prefabrication and construction in bridge engineering. Considering the structural design of steel plate composite girder bridges, which are wide but shallow in depth, their deck slabs are vulnerable to vertical impacts from vehicle loads. Structural performance may be negatively affected by excessive dynamic displacement of deck slabs. It is difficult to assess the dynamic response of the deck slabs by existing methods, since traditional specifications only use a global impact factor to describe the dynamic effect of moving vehicles on the bridge as a whole, regardless of the local dynamic effect on the deck slabs. Therefore, this study aims to assess the local dynamic effect of moving vehicles on the deck slabs of steel plate composite beam bridges using field tests and finite-element methods. A systematic approach was employed to analyze parameters influencing bridge-vehicle interaction. Additionally, an improved method was presented to calculate the local impact factor and parametric studies were discussed. The findings indicated that the local impact factor of deck slabs is significantly greater than the global impact factor. Road surface roughness is the most significant parameter affecting deck slab dynamic behavior.

Graphical abstract

Keywords

steel plate composite girder bridge / deck slab / local impact factor / field test / finite-element modeling / parametric study

1 Introduction

Rapid expansion of the highway network has driven the development of construction technology for prefabricated bridges in recent years. Modern bridge construction projects increasingly use steel–concrete lightweight composite bridges, which have advantages such as high capacity, ease of construction, and excellent ductility [1]. The composite bridge system, with its distinctive structural configuration characterized by a wide but shallow depth, is known to be susceptible to vertical vibrations that result in excessive dynamic deflection [2,3]. Hence, it has become increasingly important to study the dynamic behavior of steel–concrete composite bridge system under traffic loading conditions as modern traffic loads and speeds increase. This is a guide for the access to existing bridges and the construction of new-designed bridges.

In various design codes, “dynamic impact factors” or “dynamic load allowances” are adopted to represent the dynamic impact behavior of moving vehicles on highway bridges [4]. The specifications of each national highway bridge code provide different impact factors due to traffic conditions differences. Impact factors were determined by the length of the bridge span in the United States, New Zealand, Australia, Japan, China, and other countries previously. Some countries have revised their impact factors based on actual measurements and statistical analyses. The American Association of State Highway and Transportation Officials Load and Resistance Factor Design (AASHTO LRFD) specifications recommend a dynamic impact factor of 0.33 [5]. The Chinese bridge design code [6] employs a function of structural fundamental frequency to determine the impact factor. Nevertheless, several studies [7–11] have concluded that the actual dynamic response of moving vehicle loads cannot be accurately characterized by the simple expressions used in codes in most cases.

Dynamic impact factors specified in design codes are determined by measuring or simulating the global dynamic effect of moving vehicles on bridges. However, local dynamic effects on the primary structural elements of bridges are not well-explored. Particularly, deck slabs are a critical structural element in a steel–concrete composite bridge system. Failure to account for local dynamic effects on deck slabs can lead to the adoption of conservative design values, which can result in inadequate actual load bearing capacity. A comprehensive understanding of their local dynamic behavior under vehicle loads is essential for developing more accurate and appropriate dynamic impact factors.

Researchers [12–15] proposed a range of methodologies founded on the fundamental theory of plate vibration. These approaches were designed to examine and enhance the dynamic behavior of plates when subjected to external loads, including those frequently encountered in bridge engineering. Broquet et al. [16–18] were the first to explore the dynamic impact behavior of bridge deck slabs subjected to heavily loaded vehicles. Their investigation found that the dynamic impact factors of concrete box-girder bridge deck slabs were higher than the values prescribed in the design code at the time. The finding is noteworthy as it brings attention to the possibility of design code values being inadequate in accurately evaluating the dynamic behavior of bridge structures under realistic traffic loading conditions. Broquet et al. [19] then investigated how various parameters affect the local vibration performance of box girder deck slabs under coupled vehicle-bridge vibration conditions. During the analysis, it was determined that vehicle weight and road surface condition had a greater influence on local deck slab vibrations than any other factor. According to the report, the dynamic impact factors of concrete box girder deck slabs were between 0.3 and 0.55. Ji et al. [20,21] analyzed the local and global dynamic impact factors of a waveform steel box girder, considering parameters such as vehicle type, lane number loading, driving speed, and road surface conditions. While researchers have explored several suitable methods for calculating local impact factors, there remains a lack of comprehensive studies specifically addressing local dynamic impact factors in steel–concrete composite bridge systems.

This study aims to provide a valuable overview of assessing local and global dynamic responses for steel plate composite girder bridge and calculating local impact factors. To achieve this objective, a novel steel–concrete composite bridge system with twin girders, referred to as a steel plate composite girder bridge, located in Anhui, China, was selected and the following studies were conducted. First, field tests were conducted to measure the deflections of steel beams and deck slabs under quasi-static and dynamic loading conditions respectively with multi-point dynamic displacement optical sensors. Next, a three-dimensional (3D) finite element analysis (FEA) model of the bridge superstructure was developed and validated based on field test data. The developed FEA models enable the evaluation of bridge dynamic behavior, both local and global, under various controlled loading conditions. Considering the features of steel plate composite girder bridges and the traditional definition of impact factor, this manuscript derives the calculation method for the local impact factor appropriate for deck slabs. Then, following the comprehensive analysis of numerical simulations and experimental results, the global and local impact factors of the bridge were calculated using the traditional definition and the proposed method, respectively, and compared with the AASHTO LRFD specifications and the Chinese bridge design code. Finally, it was determined that the local impact factor is affected by a number of influences by utilizing parametric analyses, such as vehicle weight, vehicle speed, and road surface condition. The findings indicated that the local impact factor of deck slabs is significantly greater than the global impact factor. Furthermore, under several live-load conditions, the local impact factor exceeds the values specified in both Chinese and American codes. Among the various influencing parameters, road surface roughness is the most significant parameter affecting deck slab dynamic behavior.

This manuscript is organized as follows. The background of bridge is described in Section 2. Section 3 provides details of the field test. Section 4 describes the developed FEA. Section 5 accesses test and simulation results. Parameter analysis is studied in Section 6. Summary and concluding remarks in Section 7.

2 Background and bridge description

2.1 Construction description

As part of the Jinan−Qimen Highway system, the Shouchun Viaduct is located in Shouxian, Anhui, China, and was the case bridge studied in the field test. The bridge system is comprised of two separate bridges, each serving one direction of travel. As shown in Fig.1(a), for each bridge in the system, the superstructure is constructed using a continuous four-span steel plate composite girder. The overall length is 140 m, with each girder span measuring 35 m (i.e., 4 m × 35 m). As illustrated in Fig.1(b), the girder is supported at each end by two bearing pads. The bridge deck of the Shouchun Viaduct is composed of one emergency lane of 3.5 m wide, two 3.5 m wide vehicle lanes, and a curb with a width of 0.75 m. Deck parapets are installed on both sides of the bridge. A maximum speed limit of 120 km/h is achieved by the vehicle.

The steel beam is formed from Q345D I-type straight web and is 1.75 m in height, with upper and lower flanges measuring 0.8 m wide and 25 mm thick. The web has a thickness of 16 mm. The distance between the double beams is 7.23 m. Steel crossbeams are used to strengthen the transverse connection, spaced at a standard distance of 5.0 m. There are M24 high strength bolts connecting the main beam and the cross beam. Cross-sectional views of the bridge are shown in Fig.2 and Fig.3.

A structural concrete deck consists of three parts: the precast concrete slab, the longitudinal and transverse wet joints. Four precast concrete deck slabs were distributed transversely across two steel cross beams, each measuring 2.55, 3.0625, 3.0625, and 2.55 m in width. The deck slabs and steel beams act as a composite structure connected using shear connectors. Wet joints are used between the deck slabs. The longitudinal wet joints measure 0.70 m and 0.40 m wide, respectively, while the transverse wet joints are 0.3 m in width. A cross-sectional view of the deck slab shows that the standard thickness is 25 cm. The thickness of the deck slab increases to 35 cm where it contacts the steel beam supports at both ends. The positive moment slabs are made from C40 concrete, and the negative moment slabs are made from Polyvinyl Alcohol (PVA) fiber concrete. The geometric configuration of the deck slabs is shown in Fig.4. The standard cross section of the bridge is shown in Fig.5.

2.2 Cracking evaluation

Trial operation of the bridge began on December 31, 2016, following its completion on February 30, 2016. During the period of April 16 to April 21, 2020, cracks in the deck slabs and piers of bridges on the Jinan−Qimen highway system were comprehensively inspected. Fig.6 shows the cracking of deck slabs in the field. As a result of sampling observations, numerous transverse and diagonal cracks were observed at the bottom of the deck slabs, potentially compromising both the capacity and durability of the bridge. The number of precast concrete slabs with cracks was 99, representing 35% of the total number, of which five had crack widths exceeding the specification limit (> 0.2 mm). Cracking in precast concrete slabs reached a maximum width of 0.33 mm and a maximum depth of 210 mm.

In addition, it was observed that precast concrete slabs with cracks were located below both loading lanes, while no cracks existed on both sides of the flange slabs. The test report offered by the official institution indicated that the allowable bearing capacity of the deck slabs was greater than the live-load effect of the pre-designed specification [6]. The results of numerical calculations also revealed that the calculated stress of various elements was within acceptable limits. Therefore, one of the direct reasons for cracking may be that the actual live-load effect (primarily moving vehicle load impact) might exceed the considerations of the specification, which means that the deck slab design bearing capacity failed to satisfy the actual live-load effect.

3 Field test validation

3.1 Dynamic deflection measurement equipment

The aim of this research was to analyze the dynamic behavior of steel plate composite girder bridge deck slabs in service. For the purpose of establishing a monitoring mechanism for in-service behavior, a multi-point dynamic displacement optical sensor based on Digital Image Correlation (DIC) [22–24] was deployed on the superstructure to monitor structural dynamic deflection response. DIC photogrammetric instrumentation, as an effective non-destructive testing means, has the advantages of efficiency, non-contact, high accuracy and simple operation. There are several components included in the system, including Charge Coupled Device (CCD) industrial cameras industrial cameras with a 12–120 mm zoom lens, monitoring target points, a data transmission line, and a terminal station.

Field tests were conducted on the bridge span with the most severe cracking in the deck slab. Monitor target points were placed on the test bridge of the steel beam and deck slab to measure dynamic deflection. Fig.7(a) depicts the infrastructure instrumentation layout of the monitoring target points, while their detailed locations are illustrated in Fig.1(b), where target points 1 and 2 are located in the middle of the steel beam (below the emergency lane) and deck slab (middle of the single span), respectively. The CCD industrial cameras used in the study have dimensions of 50 mm × 50 mm × 50 mm and a field test focal length of 50 mm, providing a high level of accuracy with a precision of 0.001 mm. As shown in Fig.7(b), the cameras were placed under the bridge in an open and clear area. The lead wires from the cameras were connected to the terminal station computer, where camera 1 monitored the target point 1 and camera 2 monitored the target point 2, as shown in Fig.8(a). The CCD industrial cameras were placed on the ground approximately 20 m horizontally from the target points. Influenced by several obstructions, the lens of industrial CCD cameras was set at an inclined viewing angle. To eliminate slight perspective deviations, the cameras must thus be calibrated based on the 2-dimensional (2D) plane. This can calculate the perspective transformation parameters and aberration errors to enhance the measurement accuracy of the DIC method, as shown in Fig.8(b) and Fig.8(c).

3.2 Loading scenarios

To measure the actual live-load effect, in situ testing of the bridge was conducted under static and dynamic vehicle loading scenarios. The load testing involved using only one 31-t truck with four axles, which was weighed before testing. Section 4 “Finite-Element Modeling” provides detailed information regarding testing truck specifications. During the load test, one traffic lane was only restricted because of heavy traffic conditions. However, to ensure the rigor of the test procedure, the testing was chosen when no other vehicles passed. The testing vehicle was driven from one end to the other of the bridge at various predetermined speeds.

3.2.1 Quasi-static test

To simulate the general scenario of a vehicle during normal operation, the vehicle was placed along the centerline of the traffic lane. As depicted in Fig.9(a), the centerline of the test truck was approximately 0.5375 m from the west side parapet of the bridge. To ensure quasi-static conditions during the test, the driver was directed to traverse the test bridge slowly at a speed of less than 5 km/h, and then to pause for at least 1 min at a predetermined longitudinal position. This procedure was designed to induce maximum static deflection of the monitored steel beams and deck slabs. The deflection time series were recorded for target points 1 and 2 during the test. Fig.9(b) shows the driver driving the test truck safely crossing the bridge.

3.2.2 Dynamic test

For the dynamic load test, only vehicle speed was varied to investigate effects on the dynamic amplification factor of the steel plate composite girder bridge. Vehicle speed is a important parameter in determining the dynamic amplification factor [8,25–28], but its correlation with the factor remains unclear. As with the quasi-static test, the truck in the dynamic load test was positioned on the centerline of the traffic lane and driven over the test span of the bridge at three different levels (i.e., 20, 40, and 60 km/h). The speed levels were chosen for several reasons, as follows. In the pre-load test, it was found that the dynamic deflection time series at target points 1 and 2 were similar to those obtained in the quasi-static test when the vehicle traveled at speeds lower than 10 km/h. It is also difficult for the driver to control the truck at a low speed level for a long time. The truck was unable to accelerate beyond 60 km/h due to the inadequate length of the restricted traffic lanes in situ. To ensure the safety of the testing procedure, the vehicle speed for the dynamic test was limited to 20 km/h as the minimum speed, and increased in increments of 20 km/h, with the maximum speed set at 60 km/h.

4 Finite-element modeling

4.1 Steel plate composite girder bridge

A finite element (FE) model of the superstructure of the steel plate composite girder bridge is presented in this section. It was performed using ABAQUS software (version 2021). In the field test, a maximum deflection of less than 6mm was recorded, indicating that the materials behave linear elasticity. This result eliminates the demand for nonlinear geometric analysis. A detailed FE model was developed, which consisted of several physical components, including the steel beam, concrete deck slab, bearing pads, and asphalt pavement, as depicted in Fig.10. Additionally, the effect of parapets was simulated by applying pressure to the model.

In the present FE model, the simulation of steel beams was carried out using the first-order thick-shell-4 element (S4R). This type of element is a general-purpose shell element that offers 6 degrees of freedom per node, including 3 displacements and 3 rotations, applicable to thick and thin shell applications and permitting large strains. Both the C40 concrete deck slab and asphalt pavement were modeled using 8-node hexahedral solid elements (C3D8R), each node having three translation degrees of freedom.

According to the specifications of the American Concrete Institute 314-14 [29], the determination of the elastic modulus (E_c) of concrete can be achieved by employing the equation that relates E_c to f'_c, which represents concrete’s cylindrical compressive strength. E_c of asphalt pavement is derived from the modulus of fine-grained asphalt concrete at room temperature. Additionally, the bearing pads were modeled utilizing C3D8R elements, featuring an elastic modulus of 206 GPa and a Poisson ratio of 0.3.

In terms of component interaction, the asphalt pavement and the concrete deck slab were interconnected by mutual tying. As concrete deck slabs and steel beams are fixed with shear connectors, they are tied together during assembly. Since the bearings are classified into various types, the interaction properties between the superstructure and the bearings are taken into account in the finite element model by constraining or releasing the translational displacement (as shown in Fig.10), while the specific translational displacement specifications of the bearings are indicated in Fig.1(b). Tab.1 summarizes the material properties of each component in the FE model.

4.2 Vehicle

A 3D four-axle vehicle model was developed based on the test truck as depicted in Fig.11. The model was formed by the integration of rigid bodies linked together by connecting springs, axial supports, and damping systems. This configuration was adopted to accurately simulate the vehicle’s vertical, pitch, and rolling motions. To ensure coordination of vertical displacement, general contact is implemented as the contact property to simulate the interaction between the wheels and the deck slabs. A tangential friction coefficient of 0.2 is designated for tangential behavior. The boundary condition selects uniform velocity to simulate vehicle speed. Detailed mechanical and geometric parameters for the four-axle vehicle model are provided in Tab.2.

4.3 Road surface roughness

According to the 2020 version of the AASHTO LRFD specifications [4], road surface roughness constitutes a significant source of excitation for bridge vibration. It is commonly assumed that a road surface profile follows a stationary Gaussian random process with zero mean. To generate such a profile, an inverse Fourier transformation technique can be employed based on a power spectral density (PSD) function. This function can be expressed as follows:

(1)r(x)=\sum _{k=1}^N\sqrt{2\phi \left(n_k\right)\text{Δ}n}\mathrm{c}\mathrm{o}\mathrm{s}\left(2\text{π}{n}_{k}x+{\theta }_k\right),

(2)

A detailed description of this formula can be found in the related literature review [30–32]. The International Organization for Standardization (ISO) [33] has proposed a road surface roughness classification index from A (very good) to H (very poor) according to different values of \phi \left(n_0\right). As shown in Fig.12, the ISO classification of road surface roughness was used in this research. In the FE model, bridge surface conditions at different levels are simulated by modifying the nodal coordinates of the deck slab components.

5 Test and simulation results

5.1 Validation

To validate the precision of the advanced dynamic deflection measurement equipment program, an evaluation was conducted with a high-precision level gauge. Two supplementary measurement points were established at midspan, located near the positions of target points 1 and 2, but not precisely. Fig.13 presents the deflection response time series in the quasi-static test collected at these target points for moving vehicles crossing the bridge from one end to the other. The findings revealed that the maximum static deflections at Target points 1 and 2 obtained by dynamic deflection measurement instrumentation were 3.669 and 3.929 mm, respectively. These correspond to the trucks being stationary at 15.83 and 17.50 m along the longitudinal location of the bridge. The static deflections at the two temporary level points were measured at 3.654 and 3.956 mm, respectively, which demonstrated a satisfactory match with those based on photogrammetric instrumentation. In both methods, the measurement points are positioned differently, which explains the slight discrepancy.

It is evident in Fig.13 that the measured and simulated dynamic response curve shapes and peak values are closely aligned under quasi-static test conditions. It was observed that the peak values of the dynamic deflection response of the deck slab appeared prior to those of the steel beams according to the results. The reason for this is that the spacing between the cross beams is smaller than the total length of the test vehicle, which results in the vehicle load not being fully distributed on the deck slab. Additionally, piezoelectric accelerometer measurements were utilized to ascertain the first natural vibration frequency of the case bridge, which was found to be approximately 2.930 Hz. This value was observed to be in good agreement with the simulation frequency of 2.814 Hz. The first eight mode shapes and mode frequencies from the FE model simulation are summarized in Fig.14.

Shouchun Viaduct, which was built in a short time period, still has very good pavement condition. Therefore, the road surface roughness level in the FE model is defined as “very good”. Fig.15 shows the deflection measurements acquired during dynamic testing for Target points 1 and 2, as well as the related FE model analyses outcomes. Each point (or line) in these figures represents the deflection response of the corresponding target point acquired from the test (or FE model). The dynamic deflection response shows that the forced mode responses of the bridge consist of one primary harmonic and multiple frequency-multiplication steady-state harmonics under the influence of excitation load [34]. The comparison between the measured and simulated dynamic responses showed remarkable agreement, both in terms of the response curve and the peak values. This consistency validates the effectiveness of the proposed vehicle-bridge model.

5.2 Assessment of impact factors

5.2.1 Local impact factor

Structural impact amplification involves the comprehensive impact of dynamic loads (such as moving vehicles) and static loads (such as the weight of the bridge or stationary vehicles) on the bridge. The dynamic impact factor IM + 1 [35] is thus typically defined as the ratio between the maximum dynamic response caused by moving traffic and the maximum static response caused by stationary traffic. According to the traditional definition, the impact factor can be calculated as follows:

(3)IM=\frac{R_{\text{dn}}-R_{\text{sta}}}{R_{\text{sta}}}\mathrm{o}\mathrm{r}IM+1=\frac{R_{\text{dn}}}{R_{\text{sta}}},

where R_sta is the maximum static response at a particular location, and R_dyn is the maximum dynamic response at the same location. The dynamic impact factor, caculated through Eq. (3), is derived from measurements or simulations of traffic effects on the global structure of the bridge, and is thus termed global impact factors in the study.

Deflection data are a basically significant indicator of structural global behavior. It is essential to emphasize that this deflection measurement is conducted within a global coordinate system, as depicted in Fig.16(b) for \gamma and {\gamma }_1 (representing the deflection response of deck slabs AabB and steel beams A′B′, respectively). From a local perspective, as shown in Fig.16(a), a single deck slab of the steel plate composite girder bridge exhibits a deflection response {\gamma }_2 when subjected to a load. Therefore, the global deflection response \gamma of the deck slabs consists of two components, wherein the global deflection response {\gamma }_1 of the steel beam and the local deflection response {\gamma }_2 of the deck slab are included.

To accurately evaluate the dynamics of the local vibration problem for the deck slabs caused by dynamic traffic actions. Local impact factor (LIM) is proposed based on the local coordinate system of the deck slab:

(4)LIM=\left({\gamma }_{\text{dyn}}-{\gamma }_{\text{1,dyn}}\right)/\left({\gamma }_{\text{sta}}-{\gamma }_{\text{1,sta}}\right)-1,

where {\gamma }_{\text{sta}} and {\gamma }_{\text{dyn}} represent the maximum static and dynamic deflections of the deck slab, respectively. {\gamma }_{\text{1,sta}} and {\gamma }_{\text{1,dyn}} are the static and dynamic deflections of the steel beams, which need to be measured at the time of the maximum static response for the deck slab.

5.2.2 Assessment results

Based on the static and dynamic response of the bridge obtained from field testing and FE model simulation results, the global and local impact factors were calculated using the traditional definition and the proposed method, respectively, as shown in Fig.17. An assessment of the dynamic impact factor was carried out, and the results were compared against the specified values in both the AASHTO LRFD specifications and the Chinese bridge design code.

As shown in Fig.17(a), the analysis results, based on the traditional definition, indicated that the global impact factor of the steel beam rose with increasing vehicle speed. A similar tendency was observed in the global impact factor of the deck slab at speeds up to 40 km/h, beyond which it showed a continuous decrease. The findings also revealed that the maximum impact factors of target points 1 and 2 are very close under the field test conditions. Furthermore, these impact factors were found to be lower than those specified in both the previous [36] and revised [5] versions of the AASHTO LRFD, as well as the Chinese design code [6].

Fig.17(b) shows the local impact factor of target point 2 calculated using the proposed definition. The results showed that the local impact factors of deck slab had the same general trend as those of steel beam. However, the local impact factors of deck slab are significantly greater than those of steel beam. When the vehicle reaches a speed of 60 km/h, the local impact factor of deck slab exceeds the value specified in the 1996 version of the AASHTO LRFD, which is 0.2084.

6 Parametric study

Several studies have demonstrated that the impact factor is influenced by a combination of parameters [10,11]. Vehicle model, speed, and load, as well as bridge span and road surface roughness are considered. From the above field tests and FE model simulation results, it has been determined that the local impact factor of the case bridge is higher than the global impact factor, and the local impact factor is correlated with vehicle speed. As part of this section, the FE models presented in this study will be utilized to simulate the dynamic and static deflection changes of the bridge structure under various conditions. Based on the proposed method for calculating the local impact factor, the association between the local impact factor and the influencing parameters was analyzed.

6.1 Influence of vehicle speed

In field tests, the influence of vehicle speed on the impact factor has been demonstrated. Due to field testing conditions, actual vehicle speeds cannot reach the highway-specified maximum speed. Therefore, the FE models simulate the effect of different vehicle speeds on the impact factor. Fig.18 illustrates the variation in the impact factor at target points 1 and 2 at different vehicle speeds. The findings of the study indicate a nonlinear association between vehicle speed and the impact factor of the steel beam, with the obtained maximum impact factor being less than the specified values outlined in both the 1996 version of AASHTO LRFD and the Chinese design code (for convenience of description, referred to as “the two specifications”). The observed correlation is characterized by an initial increase followed by a decrease, which aligns with previous observations [37–39]. This phenomenon is attributed to the interaction between the bridge’s natural frequency and the vibration frequency of the vehicle, where a resonant frequency is achieved between the vehicle and the bridge at a velocity of 60 km/h. Other relevant literature [40] explains the phenomenon. The local impact factor of the deck slab is positive with vehicle speed. For vehicle speeds up to 120 km/h, the obtained maximum local impact factor is 0.5750, which is higher than the design values for the two specifications.

6.2 Influence of vehicle weight

It has been shown through previous research and field tests that vehicle weight is an influential parameter affecting dynamic impact. The variation in the impact factor of the case bridge for different vehicle weights is given in Fig.19, which shows the variation curves for a single truck driving at 80 km/h and under three levels of road surface roughness conditions. The results indicated that the impact factor of steel beam and deck slab decreased with increasing weight, which is consistent with previous research [41,42]. In the light-mass low moving speed region, the deck slab of the steel plate composite girder bridge exhibits weak modal interaction [43], which of the bending mode is thus more obvious, resulting in an increased impact factor. It is worth noting that when a 100-t heavy vehicle drives across the bridge in very good road surface, the obtained maximum impact factor can meet both specifications limits. However, the bridge suffers considerable deformation under heavy loads, which may compromise its safety.

6.3 Influence of road surface roughness

As a result of the long-term operation of the bridge, road surface conditions may deteriorate. Fig.20 illustrates the impact factors for target points 1 and 2 at different road surface roughness levels. Results show that the impact factors of the steel beam at the first two levels are less than the design values of the two specifications. When the road surface roughness was less than “average”, the impact factor did not satisfy the design requirements. At all five levels, the local impact factor of the deck slabs exceeded the design values of the two codes. When the road surface roughness was poor, the local impact factor reached 0.93, far exceeding the 0.33 specified in the current AASHTO LRFD [5].

7 Summary and concluding remarks

This research investigated the differences between global and local impact factors of steel–concrete composite bridge system with twin girders. The novel composite girder system constructed on the Shouchun Viaduct in China was tested under static and dynamic loading conditions to acquire adequate data for characterizing the dynamic behavior of the bridge. A multi-point dynamic displacement optical sensor was deployed on the bridge superstructure to measure dynamic deflection during loading by moving vehicles. A sophisticated full-scale 3D finite element model was developed and fine-tuned using live-load performance test data to analyze its actual performance under operation conditions. A local impact factor calculation method applicable to deck slabs is proposed. Based on a comprehensive analysis of numerical analysis and experimental results, the global and local impact factors of steel plate composite girder bridges were evaluated and compared with the provisions of the AASHTO LRFD and the Chinese bridge design code. Parametric analyses were conducted to examine the factors that affect the local impact factor. Based on the results of the research, the following conclusions can be drawn.

1) The precision of advanced dynamic deflection measurement instrumentation was validated through the comparison of mid-span deflection measurements obtained from a multi-point dynamic displacement optical sensor with those obtained from a high-precision level gauge. The findings demonstrated that there was a high level of agreement between the two methods, demonstrating the accuracy of measurement instrumentation. The simulated dynamic deflection reactions were in agreement with the measured deflection, indicating that the vehicle-bridge model can be applied to simulate other load scenarios as well.

2) Based on field tests and FE model simulations, the local impact factors of the deck slab are higher than those of the steel beam at various driving speeds. It is demonstrated that the traditional definition method underestimates the impact of moving vehicle on the deck slabs, which leads to the design bearing capacity of the Shouchun Bridge deck slab being lower than the actual live-load effect, which may be a contributing factor to the cracking observed in the deck slabs.

3) When the road surface roughness is “very good”, the impact factor of the steel beam increases with speed and then decreases, and the obtained maximum value at vehicle speed up to 60 km/h is lower than the value stated in the two specifications. The local impact factor of the deck slab is positively correlated with vehicle speed, reaching 0.575 at the maximum speed limit, which is greater than the 0.33 stated in the revised version of the AASHTO LRFD.

4) The impact factor of steel beams and deck slabs decreases with increasing weight. However, when the bridge is subjected to a heavy load, it is susceptible to considerable deformation, which poses a threat to its safety.

5) When the road surface roughness was poor, the local impact factor reached 0.93, far exceeding the 0.33 specified in the code.

Based on the above findings, it is inappropriate to use the global impact factor when assessing the bearing capacity of local components in bridge design. This will underestimate the impact effect of moving vehicles on local components, resulting in design bearing capacity failed to satisfy the actual live-load effect. In addition, road surface roughness is a significant influential parameter. After a bridge has been in operation for a long time, the road surface condition deteriorates and becomes rougher. This could increase the impact of vehicles on the bridge. To guarantee the secure functioning of the bridge, maintenance departments must prioritize reinforcing the bridge structure, implementing restrictions on vehicle load and driving speed, then use information technology and sensors to monitor bridge health. This approach to health monitoring is to develop a more accurate method of safety (condition) assessment of items including remaining lifespan and load carrying capacity than traditional strategies [44].

It is important to note that the findings presented in this manuscript are based on the specific case study bridge under investigation. However, the results provide a valuable overview of the most significant parameters and will be used for additional types of bridge research, which is the subject of ongoing research. Analysis of random traffic flows and multi-lane factors has not been explored in this research and would be material for a subsequent study.

References

[1] Deng S W, Shao X D, Zhao X D, Wang Y, Wang Y. Precast steel—UHPC lightweight composite bridge for accelerated bridge construction. Frontiers of Structural and Civil Engineering, 2021, 15(2): 364–377

[2] Tan G H, Brameld G H, Thambiratnam D P. Development of an analytical model for treating bridge-vehicle interaction. Engineering Structures, 1998, 20(1–2): 54–61

[3] Moghimi H, Ronagh H R. Impact factors for a composite steel bridge using non-linear dynamic simulation. International Journal of Impact Engineering, 2008, 35(11): 1228–1243

[4] Deng L, Cai C S. Development of dynamic impact factor for performance evaluation of existing multi-girder concrete bridges. Engineering Structures, 2010, 32(1): 21–31

[5] AASHTO. LRFD Bridge Design Specifications. Washington, D.C: AASHTO, 2020

[6] MTPRC. General Code for Design of Highway Bridges and Culverts. Beijing: Ministry of Transport of the People’s Republic of China, 2015 (in Chinese)

[7] Huang D, Wang T L, Shahawy M. Impact studies of multigirder concrete bridges. Journal of Structural Engineering, 1993, 119(8): 2387–2402

[8] Chang D, Lee H Y. Impact factors for simple-span highway girder bridges. Journal of Structural Engineering, 1994, 120(3): 704–715

[9] Liu C, Huang D, Wang T L. Analytical dynamic impact study based on correlated road roughness. Computers & Structures, 2002, 80(20-21): 1639–1650

[10] Deng L, Yu Y, Zou Q, Cai C S. State-of-the-art review of dynamic impact factors of highway bridges. Journal of Bridge Engineering, 2015, 20(5): 04014080

[11] Deng L, Wang W. Researchprogress on dynamic impact factors of highway bridges. Journal of Dynamics & Control, 2016, 14(4): 289–300

[12] Gupta R K, Traill-Nash R W. Vehicle braking on highway bridges. Journal of the Engineering Mechanics Division, 1980, 106(4): 641–658

[13] BakhtBPinjarkarS G. Dynamic testing of highway bridges––A review. Transportation Research Record: Journal of the Transportation Research Board, 1989: 1223

[14] AgarwalA CBillingJ R. Dynamic testing of the St. Vincent Street Bridge. In: Proceedings of the Annual Conference. Montreal: Canadian Society for Civil Engineering. 1990, 4, 163–181

[15] Green M F, Cebon D. Dynamic response of highway bridges to heavy vehicle load: Theory and experimental validation. Journal of Sound and Vibration, 1994, 170(1): 51–78

[16] BroquetCSchlafliMBruhwilerE. Evaluation of concrete bridge deck slabs. Swiss Society of Engineers and Architects, 1997, 123(25): 514–520 (in French)

[17] Broquet C, Bailey S F, Bruhwiler E. Dynamic amplification factors in deck slabs of concrete road bridges. Structure and Dynamics, 1999, 2: 795–799

[18] BroquetC. Dynamic Behavior of Concrete Bridge Decks Under Road Traffic Loads. Lausanne: EPFL, 1999 (in French)

[19] Broquet C, Briley S F, Fafard M. Dynamic behavior of deck slabs of concrete road bridges. Journal of Bridge Engineering, 2004, 9(2): 137–146

[20] JiWDengLHeW. Vehicle-bridge coupled vibration analysis and calculation of dynamic impact factor for the PC box-girder bridge with corrugated steel webs. Journal of Vibration Engineering, 2016, 29(6): 1041–1047 (in Chinese)

[21] JiWDengLHeW. Local and global impact factors analysis for PC box girder bridges with corrugated steel webs. Journal of Vibration and Shock, 2017, 36(8): 22–28 (in Chinese)

[22] Dai F, Lu M. Assessing the accuracy of applying photogrammetry to take geometric measurements on building products. Journal of Construction Engineering and Management, 2010, 136(2): 242–250

[23] RobsanSKyleSHarleyI. Close Range Photogrammetry: Principles, Techniques and Applications. Dunbeath: Whittles Publishing, 2011

[24] Yanushkin V N, Kolyada Y B. Complex solution of the problem of increasing the precision of linear-angular measurements by means of photoelectric methods. Measurement Techniques, 2014, 57(3): 255–259

[25] Green M F, Cebon D, Cole D J. Effects of vehiclesuspension design on dynamics of highway bridges. Journal of Structural Engineering, 1995, 121(2): 272–282

[26] Calçada R, Cunha A, Delgado R. Analysis of traffic-in-duced vibrations in a cable-stayed bridge. Part I: Experimental assessment. Journal of Bridge Engineering, 2005, 10(4): 370–385

[27] Kwasniewski L, Li H Y, Wekezer J, Malachowski J. Finite element analysis of vehicle-bridge interaction. Finite Elements in Analysis and Design, 2006, 42(11): 950–959

[28] Li H Y, Wekezer J, Kwasniewski L. Dynamic response of a highway bridge subjected to moving vehicles. Journal of Bridge Engineering, 2008, 13(5): 439–448

[29] ACI. Building Code Requirements for Structural Concrete. ACI 318-14. Farmington Hills, MI: ACI, 2014

[30] Dodds C J, Robson J D. The description of road surface roughness. Journal of Sound and Vibration, 1973, 31(2): 175–183

[31] Wang T L, Huang D Z. Cable-stayed bridge vibration due to road surface roughness. Journal of Structural Engineering, 1992, 118(5): 1354–1374

[32] Huang D Z, Wang T L. Impact analysis of cable-stayed bridges. Computers & Structures, 1992, 43(5): 897–908

[33] ISO8608. 2016 Mechanical Vibration––Road Surface Profiles––Reporting of Measured Data. Geneva: ISO, 2016

[34] Sun Z, Sun L M, Xia Y. Multi-harmonic forced vibration and resonance of simple beam bridges to moving loads. Frontiers of Structural and Civil Engineering, 2023, 17(7): 981–993

[35] Paultre P, Chaallal O, Proulx J. Bridge dynamics and dynamic amplification factors: A review of analytical and experimental findings. Canadian Journal of Civil Engineering, 1992, 19(2): 260–278

[36] AASHTO. Standard Specifications for Highway Bridges. Washington, D.C.: AASHTO, 1996

[37] HayesM DHaramisJ E I ILeskoJ J. Implementation and Non-Destructive Evaluation of Composite Structural Shapes in the Tom’s Creek Bridge. Virginia Center for Transportation Innovation and Research, 2000

[38] NeelyW D. Evaluation of the in-servic performance of the Tom’s Creek Bridge. Dissertation for the Doctoral Degree. Blacksburg, VA: Virginia Tech, 2000

[39] Li Y, Cai C S, Liu Y, Chen Y, Liu J. Dynamic analysis of a large span specially shaped hybrid girder bridge with concrete-filled steel tube arches. Engineering Structures, 2016, 106: 243–260

[40] Ma L, Zhang W, Han W S, Liu J X. Determining the dynamic amplification factor of multi-span continuous box girder bridges in highways using vehicle-bridge interaction analyses. Engineering Structures, 2019, 181: 47–59

[41] Kim S, Nowak A S. Load distribution and impact factors for I-girder bridges. Journal of Bridge Engineering, 1997, 2(3): 97–104

[42] Laman J A, Pechar J S, Boothby T E. Dynamic load allowance for through-truss bridges. Journal of Bridge Engineering, 1999, 4(4): 231–241

[43] Sun Z. Frequency comb free vibration behavior of a single-span plate pumped by low-speed moving inertial loads. International Journal of Structural Stability and Dynamics, 2022, 22(7): 2250032

[44] Miyamoto A, Kiviluoma R, Yabe A. Frontier of continuous structural health monitoring system for short & medium span bridges and condition assessment. Frontiers of Structural and Civil Engineering, 2019, 13(3): 569–604

RIGHTS & PERMISSIONS

Higher Education Press