Behavior of Lap Splice Reinforcement Bars in Light weight Concrete

Using the lightweight concrete (LWC) in the RC structures has become widespread because of its certain effect on reducing the structures self-weight. Recently Expanded Polystyrene (EPS) beads had used as a replacement of natural aggregate and one of many ways to produce LWC, that is due to its low density. By varying the EPS content within the mix, it's possible to obtain a structural EPS concrete characterized by perfect abilities of thermal as well as acoustic insulation and chemical resistance. Forces transfer between spliced bars through the surrounding concrete, therefore studying the bond between concrete and reinforcing spliced bars is necessary in order to investigate the efficiency of splicing. The objective of this research is to investigate the behavior of lapped splices of reinforcement bars within structural EPS lightweight concrete beams. The studied parameters in this research were the presence of splicing, bar size, spacing between stirrups, spliced length and the concrete cover. Deflection, cracking pattern, failure mode and longitudinal steel strain are presented and discussed in this study according to the Egyptian and American codes. Lightweight concrete exhibits good performance with splicing efficiency, which is directly proportional to the spliced length and the depth of the concrete cover, however inversely proportional to the stirrups spacing and spliced bar size.


II. EXPERIMENTAL WORK
Ten specimens of EPS reinforced concrete beam with 25 MPa compressive strength, 1800 kg/m3 density and cross section 200x300 mm with 2000mm span were tested in this study, in addition to two comparable normal weight concrete beams with the same concrete dimensions and compressive strength were constructed. Beams had three varied bar sizes, stirrups spacing, splicing lengths and depths of lower concrete covers. There were two referenced non-spliced beams, one for each concrete type. Table (I    With bar size (Stirrups spacing= 150mm), (bottom cover= 20mm) (L splice= 55Ø) and Bar size (Ø= 16mm). Bar size (Ø= 22mm).  With bar size (Ø= 10mm), (Stirrups spacing= 150mm), (L splice= 55Ø).and (Bot. cover= 10mm) (Bot. cover= 40mm).

B. Mixing and Casting
Many trial mixes were done to obtain the optimum dosage of every used material in order to achieve such two comparable mixes. After trailing and reaching the target, the final quantities required by weight for one cubic meter of fresh lightweight (LWC) and normal weight concrete (NWC) for the specimens are given in Table (II).  In order to have a homogeneous mix, the dry material {cement, silica fume, sand, coarse aggregate and the EPS beads} was blended in the mixer then the water mixed with viscocrete and gradually added. The polypropylene fibers were added after two minutes of mechanical mixing for previous contents. Cubes of 150x150x150mm and cylinders of 150mm diameter and 300mm depth were taken from each concrete mix in order to determine the compressive strength and splitting tensile strength. The specimens and the beams were cured by water sprinkling method until the date of testing.

C. Test Set up
Three vertical LVDT gages were used under beams at the mid span as well as the two thirds of the span between the two supports in order to measure the vertical deflection. Also, two electrical strain gauges were fixed on the two ends of each lap splice of the longitudinal reinforcement bars of each beam in order to measure the steel strain. The beams were tested using four point bending configuration to develop a constant moment region along the middle third of the span at which the spliced length of the bars locates as shown in Fig (14). The beams were loaded in 20 KN increments up to failure and the measurements were noted down during each loading stage.

III. EXPERIMENTAL RESULTS AND DISCUSSION
Load deflection, cracking pattern, initial stiffness (Pcr/ Δcr), Ductility (Deflection values at ultimate load (A. El-Azab (2014), ultimate and failure loads were plotted for each test specimen. The obtained results were analyzed and discussed as following.

A. Effect of tension lap splicing on the behavior the LWC and NWC beams.
Spliced and non-spliced beams were compared for both LWC and NWC in order to study the effect of tension lap splicing on the behavior of the two different concrete types. The structural details of the compared beams are shown in Fig. ( (15,16) show the load-deflection curves for the non-spliced and spliced beams respectively for the NWC and LWC. At the same load stages, LWC exhibited the values of deflection slightly greater than the NWC for both spliced and non-spliced beams. Since the reinforcement details are symmetric within the two types of beams and the ductility depends mainly on the reinforcement; the type of the used concrete did not affect the ductility, however affected the stiffness. According to the slope of the initial deflection at the linear stage of loading, NWC had an initial stiffness more than the LWC by about (107%) and (88%) for spliced and non-spliced beams respectively. Approaching the deflection values at the ultimate load stages for the two spliced beams reflects a better load transfer through the splices within them. It's observed from Fig. (17, 18) that using the splicing decreased the resulted ductility of the NWC and LWC beams by (20.3%) and (20.15%) respectively, while it decreased the initial stiffness by about (7%) and (16%) respectively; which indicated that from this study, spliced NWC beams had low reduction in the initial stiffness than LWC beams.   (19, 20) show crack, ultimate and failure loads. Since the strength of a beam depends on its ultimate load, minor differences were noted between NWC and the LWC strength for spliced and non-spliced beams. In this study, the splicing had an insignificant effect on the flexural strength for the NWC and LWC beams which decreased only by (1.15%) and (0.53%) , respectively, which indicates the efficiency of the splicing length of (55Ø) according to ECP in transferring loads within the spliced beams compared with the non-spliced one.    Fig. (5&6&7). The load-deflection curve of the tested beams is shown in Fig. (25). It's observed that increasing the spliced bar size within a beam from (10mm) to (16mm) and (22mm) decreased its ductility in both cases by about (40%), however, increased the initial stiffness of the beams by (60%) and (151%) respectively. Increasing the spliced bar size had increased the moment capacity of the beam while its shear capacity hadn't been affected. The resulted deflection of the beams at the sequent loading stages is inversely proportional to the reinforcement bar size. The failure mode was started as flexural, and then it changed to be shear-tension due to expanding the flexural-shear cracks near the ends of the splice upward as indicated from the cracking patterns and failure modes in Fig. (26, 27& 28). The failure occurred outside the splicing zone and not far from the splice ends. The flexural strength increased by about (86%) and (160%) when the bar size increased from (10mm) to (16mm) and (22mm) respectively as shown in Fig. (29).   Fig. (5, 8 & 9) show the variable stirrups spacing within the spliced beams. As the stirrups spacing reduced from (200mm) to (150mm) and (100mm) the ductility of the beam had increased by (4%) and (9%) respectively, and also the initial stiffness in the linear load stage increased by about (15%) and (44%) respectively as shown in Fig. (30). Decreasing the stirrups spacing caused a consequent decrease in the spacing between the developed cracks as shown in Fig. (31, 32 & 33). Cracks always occur adjacent to the stirrups position; thus the applied load was distributed to numerous and closed cracks which restrained their expansion. Failure modes for the three beams were nearly the same, however the beam E (stirrups spacing= 100mm) had the narrowest crack of failure. Increasing the transverse reinforcement enhances the shear capacity of the beam, but doesn't affect considerably its flexural capacity, as decreasing the stirrups spacing from (200mm) to (150mm) and (100mm) within a beam, increased its flexural strength only by (1.2%) and (4.2%) respectively respectively as shown in Fig. (34).  Fig.(5,10&11) indicate the structural details of the beams (BL, H & G) with different spliced lengths (55Ø, 45Ø & 35Ø respectively). At the same loading level, the splicing length inversely proportional to the beam deflection value. As the splicing length within the beam decreased from (55Ø) to (45Ø) and (35Ø), the ductility decreased by (14%) and (19%) respectively, also the initial stiffness at the linear stage decreased by (28%) and (51%) respectively as shown in Fig. (35). with different splicing lengths Fig.(36, 37 & 38) show the crack patterns and failure modes of the tested beams. It was noted that, for the small bar sizes, decreasing the splicing length didn't affect the failure mode; however it decreased the flexural capacity of the beam. The flexure failures occurred outside the splicing zone. As the splice length decreased from (55Ø) to (45Ø) and (35Ø), the flexural strength decreased by (1.8%) and (13%) respectively as shown in Fig.(39), this satisfied the splicing tension length (55Ø) determined by the ECP specifications.  Fig.(5, 12 & 13) indicate the structural details of the beams (I, BL & J) with different depths of concrete cover at tension zone (10, 20 & 40mm) respectively. It's observed that the ductility tends to be greater for the deeper stirrups beam, which have the smaller concrete cover. As the bottom concrete cover decreased from (40mm) to (20mm) and (10mm), i.e. increasing the depth of the stirrups, the resulted ductility increased by (1%) and (9%) respectively, and the initial stiffness of the beam decreased in both cases by about (50%) as shown in Fig. (40). It's observed from Fig. (41, 42 & 43) that the failure modes of the tested beams (I, BL & J) were almost the same outside the splicing zone and below the loading points, however the beam I (bot. cover= 10mm) exhibited a crushing of its thin concrete cover while failing. As the concrete cover depth increased from (20mm) to (40mm), the resulted flexural strength of the beam decreased by (3%) as shown in Fig.(44), because the excessive cover reduced the moment arm (d), i.e. the depth between the compression and tension forces within the beam section.    From the shown results, it is noted that, the normalized bond stress decreased by (7%) and (16%) in case of increasing the spliced bar diameter from (10mm) to (16mm) and (22mm) respectively, which agree with ACI, that the bond strength is greater for smaller bar sizes which are preferable in use. Decreasing the stirrups spacing from (200mm) to (150mm) and (100mm), increased the normalized bond strength by (25%) and (40%) respectively.That Agreed with ACI, which state that decreasing the stirrups spacing in the splicing zone, confines the spliced bars and increases the required force for the bond failure; thus increases the bond strength of the splicing. Reducing the splicing length from (55) to (45) and (35), increased the normalized bond strength by (11%) and (41%) respectively. As the concrete cover depth increased the bond strength increased, however this relation is not linear, as increasing the concrete cover at tension zone from (10mm) to (20mm) and (40mm), the normalized bond strength increased by (11%) and (17%) respectively.

IV. SPLICING BOND STRENGTH CALCULATIONS
V. CONCLUSIONS Based on the analysis of experimental test results, it can be concluded that 1. NWC is stiffer than LWC, while LWC is more ductile. The cracks and the failure of the non-spliced beams occurred at the mid span where the maximum tension zone, while in the spliced beams the failure occurred outside the splicing zone due to expanding of the cracks vertically till failure. 2. LWC has a good performance for the tension lap splice and its splicing behavior approximates that of NWC regarding the ductility and flexural strength, however differs, regarding the stiffness. 3. The splicing decreased the ductility of the NWC and LWC beams by about (20%) , while decreased the initial stiffness of the NWC and LWC beams by (7%) and (16%) respectively. 4. The splice bar size inversely proportional to the resulted ductility, while directly proportional to the stiffness and flexural capacity. As increasing the spliced bar size within a beam from (10mm) to (16mm) and (22mm) the ductility decreased by (40%) in the two cases and increased the initial stiffness and flexural strength by (60%) and (151%) and (86%) and (160%) respectively. 5. Increasing the transverse reinforcement, i.e. reducing the stirrups spacing within a beam, from (200mm) to (150mm) and (100mm) increased its resulted ductility by (4%) and (9%) respectively, and the initial stiffness by (15%) and (44%) respectively, while its flexural strength slightly increased by only (1.2%) and (4.2%) respectively. 6. Decreasing the splicing length from (55Ø) to (45Ø) and (35Ø) in the beam reinforcement, caused a decreasing its ductility by (14%) and (19%) respectively and its initial stiffness by (28%) and (51%) respectively. Also the flexure strength decreased by (1.8%) and (13%) respectively, which reflects the success of the equation developed by the ECP code determining the splicing length to (55Ø) in the case of tension splicing. 7. The bottom concrete cover depth commonly used (20mm) is the most efficient and preferable to use, providing the beam with the most accessible strength, while the extra concrete cover (40mm) did not provide an additional strength due to reducing the moment arm (d) within the beam section which consequently decreasing its flexural capacity.