Using Short Ultra-High Performance Concrete Link Slab as an Alternative to Steel Dowel Bars in Rigid Pavements

To prevent problems from occurring at the joint region in bridge construction, link slab was introduced to hide the apparent expansion joint and retain its function. This research describes experimental tests that investigate the use of an ultra-high performance concrete (UHPC), for its high sustainability properties, as a short link slab instead of apparent expansion joints and their respective steel dowels in joint plain concrete pavement (JPCP) while maintaining the expansion joint’s proper function. Similar to dowelled joints, the proposed joint distributes load between adjoining slabs while allowing for contraction and expansion of the pavement due to temperature and moisture variations. Modified cantilever tests were conducted on JPCP slab segments with regular dowelled joints and UHPC link slab joints to investigate the load-deflection response and load transfer of both joints and make comparison. The experimental results showed that UHPC link slab specimens exhibited less deflection at the face of the joint and less relative deflection between the adjoining slabs than steel dowel bar specimens. The test outcomes also showed that joint width did not have a noticeable influence on the load-deflection behavior of the UHPC link slab specimens, but it obviously affected the dowelled joints. Also, UHPC link slab specimens showed an improvement in ultimate load capacity by 23.33% and 11% compared with dowelled joints. Furthermore, the UHPC link slab joints exhibited a better load transfer between slab segments than the dowelled joints. It was concluded that even a short link slab can be superior to standard dowelled joints


Introduction
The plain concrete joint pavement JPCP is the most prominent type of concrete pavement used in airports and highways.Transverse joints have been utilized to accommodate strains produced by variations in the temperature and moisture of the concrete as well as the stoppage of construction activity on the pavement.When the joint width is less than 1 mm, it is non-dowelled; otherwise, it is dowelled.The epoxy-coated steel dowel bars provide a means for distributing loads between adjoining slabs and maintaining the horizontal and vertical level of adjacent slabs along the concrete pavement.There may be two major issues 1232 (2023) 012053 IOP Publishing doi:10.1088/1755-1315/1232/1/012053 2 with the steel dowel bars.The first issue is a looseness of the dowel, which is produced by both dowel corrosion and traffic repetitions with time, which causes a decrease in the surface area of steel dowels, resulting in joint distresses.The second issue is represented by the misalignment of the dowel, which happened as a result of incorrect dowel placement throughout the installation.As a result of the stresses restricting the slabs' movement, where the dowel deviates from the pavement's centre line either in vertical or horizontal directions, or both.These issues with the dowels can cause a lot of problems in JPCP, such as joint spalling, joint faulting, corner breaks, and transverse cracking.All of the above-mentioned issues lead to early pavement deterioration [1].GFRP is utilized instead of traditional steel dowels because its surface is very smooth and does not require lubrication to satisfy the dowel pull-out testing [1,2].It is also resistant to corrosion and has less bonding with concrete, which make the slab's movements less restricted, resulting in significantly lower locked-up stresses.GFRP dowels are around 50% more costly than standard steel dowel bars [3], although they have a lower long-term maintenance cost.However, due to their own weakness in the transverse direction, it is frequently not considered an ideal solution.Additionally, the GFRP dowel size must be increased in the flexural rigidity (EI) to be comparable to steel dowels, resulting in increased expense Research on the use of GFRP as a substitute to steel dowels as a device to transfer the load across the joints has already been done [4,5].This study recommends using short link slabs to connect adjoining slabs with no apparent expansion joints in JPCP.This method has been frequently implemented in bridges.Because the presence of transverse expansion joints in bridges is a major cause of bridge damage, it is preferred to remove these joints but still allowing the bridge to function properly by using a link slab.The UHPC materials were selected as a link slab for their high sustainability properties and their superior mechanical properties, which include high tensile strength, compressive strength, high durability, low shrinkage, high ductility, abrasion resistance, and superior corrosion [6,7,8].The first link slab was actually put into use in the state of North Carolina in 2005 [9] and was created using the design concept of Caner and Zia [10], who developed the basic design procedures that Gastal and Zia [11] had previously proposed.There has been a noticeable increase in research papers using UHPC in recent years, especially from 2015 to 2020 [12].This is because of its superior mechanical properties.A series of studies were conducted on the usage of the UHPC as a link slab in bridge expansion joints [13,14].However, only one recent study showed its appropriateness for JPCP by using link slabs with standard length (similar to the standard steel dowel length).In the current study, an experimental approach is used to investigate the effect of using short link slab on the load deflection response at the joint face, load transfer, and relative deflection of UHPC link slab joints compared with steel dowel joints.Different joint widths were used to analyse the joint's effect on the load transfer mechanism and the deflection.

Experimental Program and set-up
Two groups (each group has two specimens) of specimens were fabricated.The first group included epoxycoated steel dowel bars specimens (as a reference) and the second group contained UHPC link slab specimens.For the link slab specimens, a continuous UHPC link slab was used across the joint, while the reference specimens consisted of two nearby concrete blocks joined by a single steel dowel bar.This experiment is a simplified model that simulates what occurs on the pavement in reality.The models' dimensions were chosen to comply with the slab depth and spacing specifications of the American Association of State Highway and Transportation Officials (AASHTO) recommendations (1993) [15].The loaded and unloaded blocks were each 300 mm in width, 300 mm in length, and 200 mm in thickness.For the purpose of complying with the standard dowel bar spacing, the block width was 300 mm.According to AASHTO guidelines (1993), the steel dowel bar has a diameter of 25 mm and a length of 450 mm.It was also lightly lubricated in order to reduce bonding along the length of the dowel embedded in the concrete.
For UHPC link slabs, there are no recommendations or requirements for the dimensions of the UHPC link slabs because there has not been any actual study on the usage of link slabs on the road, Except for one study on the subject that was done by Al-Jelawy and kinaine [16], where they showed positive results for the use of link slab in roads.It was shown that deflection was lower with link slab compared to the steel dowels, and its performance was also good in terms of transferring the load across the joint to the adjacent slab.They used the assumptions that the link slab length would be preliminarily equal to the dowel bar length (450 mm) and that the link slab thickness would be half that of the slab (100 mm), based on previous studies on link slabs in expansion joints in bridge applications.The current study advances the investigations of the link slab usage in pavements to explore the effect of reduction in dimensions of link slab to reduce the associated costs.The length was reduced by 25% and the thickness was reduced by 30% compared with the study that was conducted recently [16].This leads to link slabs with dimensions of 340 mm in length, 300 mm in width, and 70 mm in thickness.The link slab width of 300 mm was selected to meet the spacing requirements between the dowels, and the thickness of 70 mm represents one-third of the slab thickness.Fig. 1 depicts the test setup.The testing machine's steel base supports the loaded block, leaving a space below the unloaded block that is equivalent to the stiff solid foundation under the loaded block so that the specimen is allowed to deform.Using an L-shaped frame, a monotonic load is applied to the loaded block and centered on the joint.As can be seen in Fig. 1, stiff plates are set over the reacting (unloaded) and loaded block surfaces, which are fastened by steel rods that are welded to the testing machine base.This prevents any horizontal, vertical, or lateral movement of the blocks above the base of the testing machine.A load cell is placed directly below the loaded block at the block's outside edge to record the applied load on the block, act as an initial support and prevent specimen tilting.This configuration follows the dowel bar testing setup outlined in the literature [17, 18, and 19], which is simple to use and closely simulating field behavior.
Linear variable differential transducers (LVDTs) were used to measure the deflection of the UHPC link slab and dowel bar at distances of 0, 25, 50, and 75 mm from the face of the joint.Slots (10 x 10) mm were created in the unloaded (reacting) block along the dowel bar centerline, and the dowel bar deflection was measured at each of these four locations.Since the deflection after 80 mm from the joint face was observed in previous studies to be minimal, no additional deflection station was made [18].The LVDTs were held in place inside the concrete slots by a specially made frame, which included a plate with holes to hold the sensors.To ensure that the sensors remained vertical during testing, these holes were accurately punctured with very little clearance.The plate was secured to the upper surface of the (reacting) unloaded block using adhesives.Two casting stages were performed for each specimen.For steel dowel bar specimens, the initial stage was to cast the unloaded block with the dowel bar accurately positioned in the center of the specimen.The second stage was to cast the loaded block.
For the specimens with UHPC link slabs, the first casting phase was to pour both conventional concrete blocks.After that, shear studs (70 mm in length and 19 mm in diameter) were installed at the middle of the region following the debonded zone (Fig. 2-c as shown in the red-colored line) to act as a connection between the UHPC and the pavement in both the loaded and unloaded blocks.The debonding zone corresponds to 50 percent of the UHPC link slab length of each block.The debonded zone is located directly underneath the UHPC link slab, which was cast in the second stage according to its mix proportions, and it was made using a duct tape starting from the face of the joint (see Fig. 2).It should be noted shear studs in such setup would restrict the horizontal movement of the joint during expansion and contraction, but it is overcome by the debonding zone which relieves the restriction and allow the UHPC to have insignificant upward movement, which was deemed acceptable when this joint was used in bridge expansion joints [13,14].

Steel Dowel Bar
According to AASHTO (1993) recommendations, the steel dowels used in this investigation had a diameter of 25 mm and a length of 458 mm.The bars in this study were made from grade 40 mild steel.Tensile strength tests were performed on three dowel specimens and the average test results were 284 MPa for yield strength and 443 MPa for tensile strength, which satisfy ASTM A615 standards [20].To protect the dowels from corrosion and reduce their bond with concrete, they were coated with a thin epoxy layer.Additionally, the steel dowels were lubricated with a thin layer of oil (SAE 10).

Normal Concrete
The test specimens were cast at the Structures laboratory at University of Al-Qadisiyah.Three concrete cylinder specimens were collected to determine the compressive strength of concrete according to ASTM C39 [21].The average compressive strength was 24 MPa at 28 days and 31 MPa on the testing day of the specimens.The following nomenclature is used to describe the test specimens: (RP) for the reactive powder concrete (RPC), i.e., UHPC, (S1) for the number of shear studs on each side of the specimens, (DB) for the dowel bar, (C) for conventional concrete, ( 10) and ( 20) for the joint width, (340) for the UHPC link slab length, and (70) for the thickness of the UHPC link slab.

UHPC Link Slab
Due to its superior mechanical characteristics and durability, the UHPC has been commonly utilized in a wide range of facilities, and studies about the UHPC performance has grown in recent years [12,[22][23][24][25][26][27].
Depending on mix proportions and the curing conditions, UHPC compressive strength when steel fiber is present is typically greater than 150 MPa [28].The average compressive strength of UHPC was 99 MPa at 28 days.On the test day, UHPC had an average compressive strength of 118 MPa.The components and mixing proportions needed to create the UHPC are listed in Table 1.Fig. 3 depicts the relative deflection between the loaded and unloaded blocks for the tested specimens.The results indicated that the UHPC link slab relative deflection, at a particular load level, is smaller than that of the steel dowels after an average deflection of about 0.1 mm.This may be due to the elimination of the apparent expansion joints and the fact that the UHPC link slab has a larger surface area than the dowelled joint, in which small contact area raises bearing stress and deflection.Furthermore, for joint widths of 10 mm and 20 mm, the ultimate relative deflection for the UHPC link slab specimens was reduced by 71.81 and 63.27 percent, respectively, when compared to the steel dowel specimens.A 20 mm joint width resulted in a very slight increase in the relative deflection of the UHPC link slab compared to a 10 mm joint width.
The results were very close to those reported by Al-Jelawy and kinaine [16] with a link slab length of 450 mm and thickness of 100 mm.Since the relative deflection was not significantly impacted by the reduction in dimensions in the current study, the proposed reduction is successful, and it leads to cost reduction.The failure modes of specimens are shown in Fig. 4. The failure mode of steel dowel bar specimens was due to concrete bearing failure as expected (see Fig. 4 (a)).UHPC link slab specimens failed by forming a single flexural crack in the UHPC that propagated to the top surface as shown in Fig. 4 (b).

Load Deflection Behavior
Fig. 5 describes the load-deflection behavior of the steel dowel and the UHPC link slab specimens at the face of the joint.In comparison to the dowel bar, which has a lower surface area than the link slab, the applied load is spread over a wider area in the UHPC link slab, resulting in less deflection than the dowels by 36% and 82.22% respectively for 10 mm and 20 mm joint width.The UHPC link slab specimens exhibited an enhanced ultimate load as compared to dowel bar specimens by 23.33% and 11% for joint widths of 10 mm and 20 mm, respectively, as shown in Figs.5(a) and (b).Load enhancement was lower than that in the case with the link slab length of 450 mm [15], which is to be expected given the reduction in link slab length to 340 mm.Figs. 5 (c) and (d) depict the effect of joint width on deflection, where the deflection did not sound to be affected by the joint width of UHPC link slab specimens.However, the deflection was 18.87% more for joint width of 20 mm than for joint width of 10 mm for steel dowel bar specimens.Table 2 displays the total load and joint face deflection for all specimens.In JPCP, the ideal state for the transfer of load across the joints is 100%.However, under normal circumstances, this is challenging to achieve.Load transfer efficiency (LTE) can be obtained using Equation (1).The LTE for the study specimens is shown in Fig. 6.The x-axis represents the total applied load, while y-axis represents LTE.The steel dowel specimens are represented by the first number in the xaxis, whereas the link slab specimens are represented by the second number.The LTE for the link slab specimens is about the same as or slightly higher than that of the steel dowel specimens, which was approximately (70-90) % on average.A rise in the load carried across the joints was obtained by the UHPC link slab specimens, where their larger surface area and the existence of shear studs facilitated that.The LTE did not seem to be affected by joint width as seen in Figs. 6 (a) and (b).LTE was not affected by the length reduction (340 mm) compared with what was reported previously [16] for link slab length of 450 mm as the current results were somewhat similar.
Where; PT and P represented the total transferred load and the total applied load, respectively

Conclusions
An experimental investigation using modified cantilever tests was carried out to explore the load deflection response of steel dowel and UHPC link slab that link JPCP specimens for two different joint widths, as well as to compare load transfer efficiency and the relative deflection between these two types of joints.Two of the test specimens had steel dowels, while the other two had UHPC link slabs.
The experimental results showed that UHPC link slab specimens exhibited reduced deflection at the joint face by 36% and 82.22% compared with dowelled joints and reduced relative deflection between the adjoining slabs by 71.81 % and 63.27% compared with steel dowel bar specimens for joint widths of 10 mm and 20 mm, respectively.The test outcomes also showed that the load-deflection response of the UHPC link slab specimens was not affected by the joint width, but it was clearly affected for the dowelled joints.Also, UHPC link slab specimens showed an improvement in ultimate load capacity by 23.33% and 11% for joint width of 10 mm and 20 mm, compared to dowelled joints.Furthermore, both joint types exhibited load transfer efficiency (LTE) of at least (70-90) % with slightly better performance of UHPC link slab joints compared with steel dowel bar joints.Short link slab was found to be superior to standard dowelled joints.

Figure 1 .
Figure 1.The test setup: (a-b) test setup schematic and in the laboratory for dowelled-joint specimens; (cd) Test setup schematic and in the laboratory for the UHPC link slab specimens.

Figure 3 .
Figure 3. Dowel bar and UHPC link slab relative deflection

Figure 4 .Figure 5 .
Figure 4. Failure Mode of the steel dowel and UHPC link slab specimens

Figure 6 .
Figure 6.Load transfer efficiency (LTE) for steel dowel and link slab specimens

Table 1 .
Proportional and Materials of UHPC

Table 2 .
Deflection and total load for UHPC link slab and the steel dowels specimens