Introduction
There are two types of girders that are widely used in long-span bridges, i.e., orthotropic steel girders and steel-concrete composite girders. However, the two types of girders have some drawbacks. For example, the orthotropic steel decks (OSDs) are vulnerable to fatigue cracking under cyclic traffic loads; additionally, the asphalt overlays on top of the OSDs are prone to premature deteriorations such as cracking, shoving, etc. Regarding the steel-concrete composite girders, the main concerns are that: 1) the concrete deck is usually 250–400 mm thick and is thus heavy, making it uneconomical for long-span bridges (e.g., over 600 m); 2) since normal concrete has low tensile strength, the concrete deck usually develops multiple cracks at the negative bending moment zones, which lead to the reduction of the stiffness of the girder and may even influence the durability of a bridge.
To overcome the aforementioned drawbacks and thus to improve the serviceability for the two types of girders, the research group at Hunan University proposes three types of novel composite girders consisting of steel and UHPC, in which UHPC denotes ultra-high performance concrete. UHPC has excellent mechanical properties, including high strength, low creep coefficient, low risk of cracking, and impressive durability. The proposed steel-UHPC lightweight composite bridge girders include the following three types: (1) Type-1 girder, a composite girder composed of a conventional OSD and a thin UHPC layer; (2) Type-2 girder, a composite girder composed of a refined open-ribbed OSD and a thin UHPC layer; (3) Type-3 girder, a composite girder composed of a steel beam and a UHPC waffle deck.
With the advantages of both steel and UHPC, the steel-UHPC composite girders are free from premature deteriorations such as fatigue cracking in OSDs and damages in asphalt overlays; in addition, the cracking resistance of the UHPC layer (or deck) in the negative moment zones is significantly improved. All these advantages make them a competitive solution to develop efficient and durable long-span bridges.
To reveal the basic static and fatigue performance of the three types of steel-UHPC composite girders, serial theoretical and experimental studies have been performed by the research group at Hunan University. This paper outlines these studies which illustrate the excellent performance of the steel-UHPC composite girders.
TYPE-1 steel-UHPC composite girder
Configuration characteristics and advantages
As shown in Fig. 1, a typical Type-1 girder consists of a conventional OSD and a thin UHPC layer; the two components are connected by stud shear connectors. The connectors usually have dimensions of 13 mm × 35 mm (diameter × length). This type of composite girder is suitable for steel bridges adopting conventional U-ribbed OSDs, including both existing and newly constructed bridges.
The UHPC involved in the Type-1 girder is compactly reinforced with steel reinforcement bars. In common cases, conventional UHPC has a tensile strength of 8–15 MPa [
1,
2], while the compactly reinforced UHPC is revealed to have a tensile cracking strength of 42.7 MPa [
3].
Compared to conventional OSDs, the Type-1 girder has the following advantages:
1) The stiffness of the OSD is improved because of the presence of the UHPC layer. As a consequence, the vehicle-induced stresses in the OSD are significantly reduced, indicating decreased fatigue cracking risk in the OSD.
2) The UHPC layer can have 100-year service life because of its excellent mechanical behaviors and impressive durability. As a consequence, the life-long cost of the Type-1 girder is much lower. A cost comparison is made based on the Humen Bridge, a suspension bridge in China. The bridge is 888 m long and 30 m wide, with a resultant overlay area of 888 × 30= 26640 m2. In China, long-span OSD bridges usually adopt epoxy asphalt as their surfacing overlays. The unit cost of the epoxy asphalt is 1600 RMB/m2, and the time interval for overlay replacement is in general 8 years. While for the Type-1 girder, the unit cost of the UHPC and the asphalt overlay is 2000 RMB/m2 and 80 RMB/m2, respectively. The UHPC layer needs no replacement during the service life expectancy; on the contrary, the asphalt overlay has a service life of 8 years. Based on the above data and assumptions, the cost at the end of the service life expectancy (100 years) is calculated. It is revealed that the cost of the Type-1 girder is 79 million RMB, only about 15% of the cost of the normal OSD with an epoxy asphalt overlay (511 million).
3) Because that the UHPC layer is in general only 35–60 mm thick, the Type-1 girder is light and is thus especially suitable for long-span bridges.
Static performance
Behavior of shear connectors
As mentioned earlier, the UHPC layer in the Type-1 girder is thin. Thus it is required that the length of the headed studs should be controlled. Generally, the studs used in the Type-1 girder have a height-to-diameter ratio of 35 mm/13 mm= 2.7. To reveal whether the short headed studs can develop their full strengths in UHPC, push-out tests were performed. The setup of the push-out tests is shown in Fig. 2(a), and the failure model of a specimen is shown in Fig. 2(b).
Figure 2(b) indicates that the failure model of the specimens was that the studs were sheared off from the steel plate, and the UHPC layer remained intact with no cracks developed. Thus, the push-out tests revealed that the studs can develop full strengths even though they have a low aspect ratio of 2.7.
Behavior of composite beam specimen
Within a Type-1 girder, the UHPC layer usually develops flexural tensile stresses at the negative bending moment zones such as at the diaphragm sections. To reveal the behavior of the Type-1 girder in resisting such a negative bending moment, a static load test was performed on a full-scale composite beam specimen, as shown in Fig. 3(a). The specimen is composed of a longitudinal OSD strip and a 45-mm-thick UHPC layer. A load was applied near the free end so as to produce a negative bending moment in the specimen.
According to the test results, when the bottom flange of the OSD began to yield due to excessive compression, the UHPC layer did not develop visible cracks. When the UHPC layer developed an initial 0.05-mm-widecrack, more portions of the bottom of the OSD yielded. And when the load reached the peak value, local buckling was apparent at the bottom flange of the OSD (Fig. 3(b)). The test indicated that the OSD failed prior to the UHPC layer.
Fatigue performance
Influence of UHPC layer on OSD details
To quantitatively evaluate the influence of a UHPC layer on the fatigue performance of an OSD, comprehensive finite element (FE) analysis was performed. The analysis was based on an example steel bridge in China, the Humen Bridge.
The Humen Bridge was constructed in 1997. After decades of exposure to heavily-loaded traffic, the Humen Bridge has developed premature deteriorations in its deck system, including fatigue cracks in the OSD and cracking and shoving in the asphalt overlay. To address these issues, the research group has proposed a Type-1 girder scheme for the Humen Bridge. The proposed Type-1 girder scheme consists of a 45-mm-thick UHPC layer and a 25-mm-thick asphalt overlay.
Two local FE models were built by using the Ansys software, among which one modeled a normal OSD and the other modeled a Type-1 girder. Within the FE models, the OSD, the UHPC layer, and the studs were modeled via shell elements, solid elements, and beam elements, respectively. Figure 4 shows the analysis results as well as six typical fatigue-prone details of interest.
The analysis results shown in Fig. 4 reflect that the vehicle-induced stress ranges of the OSD in the Type-1 girder are 21%–82% lower than those in a normal OSD. Figure 4 also shows that some fatigue-prone details have their maximum stress ranges less than the corresponding constant-amplitude fatigue limits (CAFLs): the rib-to-deck welded joints, the free edge of the cutouts, and the butt joints of the ribs, indicating infinite fatigue life of these fatigue-prone details theoretically. However, the stress ranges at the rib-to-diaphragm welded joints are still larger than the CAFLs.
Behavior under negative bending moment
A fatigue test was performed on a full-scale steel-UHPC composite beam specimen to reveal the fatigue behavior of the UHPC layer under negative bending [
4]. The setup of the fatigue test is shown in Fig. 5. In the test, the fatigue load produced a tensile stress range of 9.8–24.3 MPa in the UHPC layer, for 3.1 million cycles.
Considering that impulse stress ranges are usually used to describe the fatigue behavior for a material, the applied stress range in the test was converted to an impulse stress range of 0–21.3 MPa based on the S-N curve of concrete. According to the test results, the UHPC layer developed no fatigue cracks after experiencing 3.1 million cycles of loading.
Applications
By 2019, the Type-1 girder has been applied to over 30 real bridges in China (Table 1), which cover the four basic bridge patterns, i.e., girder bridge, arch bridge, cable-stayed bridge, and suspension bridge. Some relevant photographs are shown in Fig. 6.
The Mafang Bridge is the first pilot project among the 14 bridges. It is a 14-span simply-supported box-girder bridge constructed in 1984, with each span being 64 m long. The deck of the bridge is an open-ribbed OSD.
The asphalt overlay of the Mafang Bridge has deteriorated seriously and several rounds of retrofit have been implemented. In 2011, another round of major retrofit was undertaken to replace the deteriorated asphalt overlay. AType-1 girder scheme was applied to the 11th span of the bridge as an alternative scheme. For comparison, four asphalt surfacing schemes were applied to the other 13 spans of the bridge.
To date, the Type-1 girder scheme of the Mafang Bridge (on the 11th span) has been in service for over 5 years, and has been tested three times. The tests revealed that the 11th span of the bridge developed no obvious deterioration. By contrast, the four asphalt surfacing schemes applied on the other 13 spans of the bridge have developed severe deterioration such as cracking, pot holes, etc. [
5].
TYPE-2 steel-UHPC composite girder
Configuration characteristics and advantages
Figure 7 shows a typical cross section of the Type-2 girder, which has the following characteristics compared to the Type-1 girder (Fig. 1): (1) the bulb flat open ribs; (2) the apple-shaped cutouts on the diaphragms. This type of composite girder is suitable for newly constructed steel bridges.
It is worth mentioning that because of the contribution of the UHPC layer to the overall stiffness of the bridge deck, the spacing of the bulb flat open ribs can be increased to 400–500 mm, a value significantly greater than the rib spacing unconventional open-ribbed OSDs (300 mm in general).
The Type-2 girder has the following advantages over the Type-1 girder: 1) lower fatigue cracking risk because of its simpler configuration; 2) easier to weld the ribs on site because of its better accessibility.
Static performance
To reveal the behavior of the Type-2 composite girder in resisting negative bending moment, a static load test was undertaken based on a full-scale strip specimen, as demonstrated in Fig. 8.
In the test, when the load reached 166.5 kN, the tensile strain of the UHPC layer reached 834 με and the corresponding tensile stress reached 35.5 MPa, but no visible cracks developed in the UHPC layer; when the load was increased to 179.0 kN, the first visible crack appeared on the top of the UHPC layer, with the maximum cracking width being 0.05 mm.
Fatigue performance
A FE analysis was first carried out to reveal the influence of the UHPC layer on the fatigue life of the open-ribbed OSD (Fig. 9(a)), and then a fatigue test was performed for a full-scale specimen (Fig. 9(b)).
The FE analysis was based on the second Dongting Lake Bridge, a 1480 m truss-girder suspension bridge in China, which is currently under construction. The analysis results (Fig. 9(a)) indicate that in the Type-2 girder, the vehicle-induced stress ranges in all of the six fatigue-prone details (Fig. 4(b)) are below their constant-amplitude fatigue limits. Thus, the fatigue life of these fatigue-prone details should be infinite theoretically.
In the fatigue test (Fig. 9(b)), the most unfavorable fatigue-prone detail was revealed as the free edge of the cutouts on the diaphragms. The fatigue test indicated that the specimen developed no fatigue cracks after experiencing 2.5 million cycles of loading. Considering that the free edge of the cutouts on the diaphragms had a maximum tensile stress range of 90.6 MPa in the test, a value much greater than the maximum stress range of the detail under design loads (49.8 MPa), the service life of the Type-2 girder should exceed 2 million cycles, which is required by the Eurocode [
6].
TYPE-3 steel-UHPC composite girder
Configuration characteristics and advantages
Figure 10 presents a schematic of the Type-3 girder. A typical Type-3 girder consists of a steel beam and a UHPC waffle deck, in which the steel beam can be either I-shaped or box-shaped.
By comparing the UHPC waffle deck to the conventional concrete deck, the following advantages are remarkable:1) The UHPC waffle deck is 50%–60% thinner and is therefore about 40% lighter, making it applicable to bridges spanning over 1000 m. In comparison, conventional concrete decks are only applicable to bridges with a span length of ≤ 400–600 m; 2) Having high tensile strength, the UHPC waffle deck outperforms conventional concrete decks in resisting cracking; 3) Prestressing tendons can be eliminated in the UHPC waffle deck because of the excellent mechanical behaviors of UHPC, which simplifies the constructional processes significantly.
Static performance
Finite element analysis was performed based on an example bridge, the Shengtian Bridge. The Shengtian Bridge is a cable-stayed bridge, and it has a main span length of 450 m. A Type-3 girder was proposed for the Shengtian Bridge, as shown in Fig. 11. The Type-3 composite girder is especially suitable for long-span cable-stayed bridges.
FE analysis was performed to explore the stress states of the UHPC waffle deck under design loads. Two FE models were built, i.e., a global model reflecting the overall behavior of the bridge and a local model reflecting the behavior of the deck. The analysis results from the two models were superimposed, and the calculation results indicated that the combined peak tensile stress in the UHPC waffle deck is 12.5 MPa.
To verify whether the UHPC waffle deck is able to resist such a high tensile stress, an experimental test was performed on a longitudinal full-scale strip specimen (Fig. 12).The specimen was subjected to positive bending moment.
The test revealed that when the tensile stress reached 15.1 MPa, the specimen developed a first crack and the crack width was 0.03 mm. Thus, by comparing the cracking strength in the test to the tensile stress in design, it can be predicted that the UHPC waffle deck should be safe in service.
Conclusions
The research group at Hunan University proposes three types of lightweight composite bridge girders based on the UHPC material, which has excellent mechanical behaviors and durability. A series of theoretical and experimental studies has been undertaken to reveal the fundamental performance of these composite girders. Based on the current investigations, the following summary and conclusions can be made:
1) The Type-1 composite girder consists of a conventional OSD and a thin UHPC layer. FE analysis has shown that due to the presence of the UHPC layer, the stress ranges caused by cyclic traffic at the following fatigue-prone details of the OSD were below the constant-amplitude fatigue limits: the rib-to-deck welds, the free edge of the cutouts on diaphragms, and the butt joints in ribs. The observation simply that these details should have infinite fatigue life theoretically. Experimental tests, including both the static and fatigue tests, were performed and the results showed that both the headed studs and UHPC layer can meet the design requirements.
2) The Type-2 girder consists of a refined open-ribbed OSD and a thin UHPC layer. FE analysis has shown that all of the typical fatigue-prone details of the OSD were found to have stress ranges less than the constant-amplitude fatigue limits, indicating infinite fatigue life of these details theoretically. The excellent static and fatigue performances of the Type-2 girder are also verified by experimental tests.
3) The Type-3 girder consists of a steel beam and a UHPC waffle deck. The stress state of the UHPC waffle deck was revealed by FE analysis based on an example bridge; then a load test was undertaken for a longitudinal strip specimen. The tensile cracking strength of the UHPC obtained in the test was compared to the maximum tensile stress of the waffle deck caused by design loads, and the comparison indicated that the tensile cracking strength of the UHPC exceeded the maximum stresses under design loads. Thus, the waffle deck can meet the requirement under the design loads.
Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature
PDF
(16459KB)
