Strengthening of hollow core precast prestressed reinforced concrete slabs using different techniques

Abstract

Hollow-core precast prestressed reinforced concrete (HCPPRC) slabs have been widely available since the 1980s in different structures due to many advantages. Since no shear reinforcement is arranged in HCPPRC slabs before casting, concrete itself provides the shear resistance, especially in the zones close to the supports, which are the most popular regions for shear failure. This experimental study aims to investigate some shear strengthening techniques for HCPPRC slabs, including using carbon fiber reinforced polymer (CFRP) sheets, concrete topping, filling 50% of the slab voids in shear zones with concrete, filling all shear zone voids with concrete, steel through anchors in shear zones, and mixing between filling 50% of the slab shear zone voids in addition to using steel anchors. The experimental program consists of two control specimens and six shear-strengthened specimens that were tested using two-line loads up to failure. Cracking load, cracking deflection, ultimate load, maximum deflection, and failure mode are recorded and discussed for all specimens. The results show that filling 50% and 100% of shear voids with concrete increases slab concrete strength by about 79% and 61.45%, respectively, while using steel through anchors decreases slab concrete strength. Moreover, filling 50% of the shear voids is the best technique for strengthening the shear capacity of HCPPRC slabs due to increasing the cross-section area of the slab’s shear strength. In contrast, using screws with nuts only is not preferred for shear strengthening of HCPPRC slabs.

Keywords:

• Prestressed
• precast
• hollow core
• shear
• strengthening
• voids

Reviewing Editor:

• Sanjay Kumar Shukla, Edith Cowan University, Australia

Subjects:

• Mechanical Engineering
• Testing
• Mechanical Engineering Design
• Structural Engineering

1. Introduction

Hollow-core precast prestressed reinforced concrete (HCPPRC) slabs have been widely used in many civil engineering structures, including commercial, residential, and parking structures. Slab systems made of concrete hollow-core units are used because of their high quality, lower amount of used materials, and capability of spanning long distances with relatively small depth. Hollow-core units contain voids that run continuously along their length, which aim to reduce the dead weight of the construction and the cost of materials. Hollow core slab (HCS) has been a hot study area for many years as a result of the aforementioned considerations in order to maximize its flexural or shear capacities (Chen et al., 2023; Ferreira et al., 2021; Foubert et al., 2016; Zając et al., 2021; Zavalis et al., 2022).

Most HCS defects result from a shortage of shear capacity due to the large proportion of voids in the cross-section without shear reinforcement, which leads to web shear cracks and failure, especially in the support region. However, experimental and numerical works available in the literature (Araujo et al., 2011; Walraven & Mercx, 1983) showed that HCSs could be characterized into four different failure modes, i.e., anchorage, flexure, shear-tension, and flexural shear, as shown in Figure 1. There are a lot of factors that affect HCS shear capacity or behaviour (El-Arab, 2017; Kankeri & Prakash, 2017; Pinheiro et al., 2023), such as cross-section, span-to-depth ratio, and the transmission length in pre-stressed concrete HCS. Brunesi et al. (2015) studied the effect of cross-section features on the shear behavior and capacity of HCS. The results showed that the brittle web-shear failure mechanism experienced was proven to be controlled by hollow core shapes and the related non-circularity of the voids. Also, Elliott (2014) studied the shear capacity and the transmission length in pre-stressed concrete HCS, which recommended reducing the transmission zone around voids to increase shear capacity and improve the failure criteria at voids in some cases. Some studies comparing traditional codes (Eurocode (EC2) (ACI Committee, 2005) and American Concrete Institute (ACI) (British Standard, 2004)) were performed. It has been determined that the method for shear resistance calculation given in Eurocode 2 appears to work well for slabs with circular voids, but for slabs with flat webs and noncircular voids, it gives nonconservative results (Pajari, 2005).

Figure 1. Failure modes near supports of HCS.

Furthermore, concerns have been raised over the lack of HCS shear resistances, especially those at the end regions, which aim to maximize its shear capacity through several techniques. Some works conducted theoretical and experimental studies of structural topping as a type of strengthening technique and its effects on HCS shear capacity and the adequacy or bond capacity of the interfaces. It has been observed that concrete topping could increase the HCS’s shear capacity by about 35%. So, the bond between the topping and the original slab is a suitable technique for retrofitting HCS, and the way the topping interacts with the original slab is acceptable behavior (Girhammar & Pajari, 2008; Pajari, 2009). Another technique of using fiber-reinforced concrete was studied to improve the shear performance of HCS and achieve more ductile performance. The use of fibers is a probable method to improve shear collapse because fibers are able to increase the strength of the element to its full flexural capacity (Cuenca & Serna, 2013; Mahmoud et al., 2017). Also, some works studied the effect of shear reinforcement (steel bars or CFRB sheets) in HCS, which showed that hollow-core elements could be used in most applications without the need for shear reinforcement, especially those that were deeper than 12.5 in. (320 mm) (Elgabbas et al., 2010; Hawkins & Ghosh, 2006; Li et al., 2018; Wu, 2015). Other works addressed similar problems of strengthening (El-Feky et al., 2023; Fayed et al., 2023; Fujikura et al., 2021; Nasery et al., 2020a, 2020b; Nasery, Hüsem, Okur, Altunışık, & Nasery, 2020; Shehab et al., 2023).

Based on the literature, it can be seen that although some strengthening techniques of hollow-core elements show their efficiency in shear resistances, there is a need to study more shear strengthening techniques that can help in improving the shear capacity of HCPPRC slabs. The main goal of this study is to shed some light on the effects of several strengthening techniques on the shear behaviour of HCPPRC slabs. These techniques included using carbon fiber reinforced polymer (CFRP) sheets, concrete topping, filling fifty percent of the slab voids in shear zones with concrete, filling all shear zone voids with concrete, steel through anchors in shear zones, and mixing between filling fifty percent of shear voids in addition to steel anchor. Experimental results such as failure mode, cracking load, and load vs. displacement relations were recorded and discussed for all specimens. The results and recommendations from this work may be useful for researchers working on strengthening the shear capacities of HCPPRC slabs.

2. Experimental work and strengthening methods

The experimental program consisted of testing eight HCPPRC slab specimens with overall dimensions of 2250 × 1200 × 300 mm (length, width, and thickness) and clear spans of 2050 mm. The slabs were produced by the Arab Contractors Company. All of these slabs were reinforced with four wires of 3/8” diameter (7 strands). The prestressing steel reinforcement is classified as a low relaxation strand with ultimate strength and strain equal to 1860 MPa and 3.50%, respectively. The clear concrete cover of the wires was maintained at 60 mm measured from the slab soffit, as shown in Figure 2.

Figure 2. Cross section of the sample.

The specimens were cast with a nominal characteristic concrete cubic compressive strength after 28 days equal to 40 MPa. The slabs were demolded after 48 h of casting and covered with wet burlap for 7 days, as shown in Figure 3. For control purposes, the concrete cubes were taken and tested by the Arab Contractors Company. The experimental program was conducted at the Housing and Building National Research Center (HBRC), Egypt.

Figure 3. Covering the sample with wet burlap for 7 days.

The strengthening techniques for s1, s2, s3, s4, s5, s6, s7, and s8 were as shown. S1 was the first control slab without any strengthening. It was reinforced with four strands with a diameter of 3/8 inch. The cube concrete strength of all specimens was 40 N/mm², as shown in Figure 4.

Figure 4. During testing the first control slab S1, without strengthening.

The slab S2 was the control specimen number (2), which was reinforced with four steel plates in the moment zone. The steel plate yield stress (fy) was 240 MPa. The steel plate dimensions were 2000 × 150 × 4 mm (length, width, and thickness). The four steel plates were used to increase the slab’s moment capacity. This is mainly to guarantee that the specimens will fail due to shear failure instead of flexural failure. The surface of the slab was cleaned and moistened. The steel plates were installed using high-strength epoxy resin adhesive; this is mainly to guarantee a full bond between the steel plates and the concrete surface. One mechanical bolt of 10 mm diameter was used to install the steel plates at both ends of the specimens, as shown in Figure 5.

Figure 5. Strengthening the second control slab S2.

The slab S3 was strengthened with four steel plates (Fy = 240 Mpa) of dimensions 2000 × 150 × 4 mm (length, width, and thickness), in addition to 32 steel screws with nuts in four rows at each side; each row contained four steel screws. The steel screws were 12 mm in diameter and of grade 10.9. The spacing between the steel screws in the longitudinal direction of the slabs is 150 mm for 600 mm (twice the specimen thickness) from each edge. It was repaired around the steel screws by using high-strength epoxy resin adhesive, as shown in Figure 6.

Figure 6. Strengthening the third slab S3.

The slab S4 was strengthened as specimen S2 in addition to 85-mm concrete topping of strength 30 MPa, in the longitudinal direction of the slabs for 880 mm from each support. Percentage of increase in total thickness/original thickness = 28.33%. There were two rows of steel bars of diameter 12-mm in L shape (fy = 360 MPa), which were distributed at 300-mm. Each row contained three anchors on each side. It was installed with a non-shrink epoxy adhesive mortar. It was used as a high-quality adhesive for mortar and concrete in the casting process of the concrete topping, as shown in Figure 7.

Figure 7. Strengthening the slab S4.

The slab S5 was strengthened by filling four voids of concrete of strength 30 MPa in the longitudinal direction of the slabs for 880 mm from each support. Using a concrete saw, cut grooves approximately 40 mm wide and 40 mm deep. The grooves were cut at the compression flange for four voids (about 50% of the slab voids). It was used as a non-shrink epoxy adhesive mortar in the casting process of concrete filling, as shown in Figure 8.

Figure 8. Strengthening the slab S5.

The slab S6 was strengthened by filling four voids (about 50% of the slab voids) with concrete of compressive strength 30 MPa, in the longitudinal direction of the slabs for 880 mm from each support, in addition to 32 screws with nuts in four rows at each side; each row contains four screws. It was used as a high-quality adhesive for mortar and concrete in the casting process of concrete filling, as shown in Figure 9.

Figure 9. Strengthening the slab S6.

The slab S7 was strengthened by a carbon fiber reinforced polymer (CFRP) sheet of width 150 mm, thickness 0.167 mm, and length 1400 mm, which was raped as stirrups with high-strength epoxy resin adhesive. CFRP properties were: area weight was 300 g/m² ± 15 g/m², fiber density was 1.79 g/cm³, tensile strength was 3900 N/mm², tensile E-modulus was 230,000 N/mm², ultimate load was 480 kN/m width per layer, and tensile E-modulus was 30 kN/mm². The slab was punched with a 20-mm punch using the halo. The surface was leveled with a high-strength epoxy resin adhesive. The CFRP stirrups were put on 4 rows, and each row contained 2 stirrups, as shown in Figure 10.

Figure 10. Strengthening the slab S7.

The slab S8 was strengthened by filling all the slab voids (100% of the slab voids) with concrete of compressive strength 30 MPa in the longitudinal direction of the slab for 880 mm from each support. Using a concrete saw, cut grooves approximately 40 mm wide and 40 mm deep. The grooves were cut at the compression flange for all voids. It was used as a high-quality adhesive for mortar and concrete in the casting process of concrete filling, as shown in Figure 11.

Figure 11. Strengthening the slab S8.

3. Test setup

All slabs were supported on joint A, which was a hinged support, while joint B was a roller support. The live photo and schematic diagram for the test set-up and instrumentation used for monitoring the shear behavior of the specimens are shown in Figures 12 and 13, respectively. The slabs were tested in shear using a hydraulic jack of 500 kN capacity that distributed the load using a rigid steel beam on two vertical loads located at 750 mm from each support. The applied load was measured using a 1000-kN load cell, which was installed between the hydraulic jack and the distributor steel beam. The rate of loading was 0.43 kN/s. Three linear variable displacement transducers (LVDT₁ to LVDT₃) were put under the points of loading and the load cell, which were used to measure the vertical upward and downward deflection during testing. The compressive strains of concrete were measured using PI gauges installed in the middle of the span at the top of the specimen. Inclined PI-gauges and LVDT transducers were used to measure the induced strains in the inclined strut composed of the loading scheme. The data was recorded automatically by using the data logger system during testing the specimens, as shown in Figure 12.

Figure 12. Live photo for the test set up.

Figure 13. Schematic diagram for the test set-up.

4. Results and discussions

Table 1 presents the measured data for the eight slabs, where flexure and shear failure were observed. Maximum loads, crack loads, maximum deflections, and failure modes for all specimens are presented.

Table 1. Experimental test results of the tested slabs.

Slab no	Δ crack(mm)	P crack (KN)	Δ maximum(mm)	P maximum (KN)	Mode of failure
S. no (1)	2.8	190	61.2	271.4	Flexure and shear failure
S. no (2)	1.2	50	9.06	262	Shear failure
S. no (3)	5.4	41.9	8.75	41.9	Shear failure
S. no (4)	2.8	384	20	384	Shear failure
S. no (5)	3.27	350	11.12	469	Shear failure
S. no (6)	3.67	300	10.91	420	Shear failure
S. no (7)	2.4	141	7.4	305	Shear failure
S. no (8)	2.63	334	15.21	423	Shear failure

4.1. Failure modes

The first crack was observed at specimen S1 at the midspan and at a load level equal to 190 kN. This was due to flexure failure. The second crack appeared at a load equal to 220 kN. At this load level, the compression failure at the top slab fiber started to fail. At a load equal to 271.4 kN, flexural failure occurred. A slippage of the pre-stressing strands occurred at the failure stage. An inclined shear crack formed at the failure stage, with an angle equal to 41 degrees. It should be noted that the failure mode was a combination of the flexural and the shear modes, as shown in Figure 14. The strain at the compression strut where the PI gauges were installed at the shear zone was -0.01%. However, the strain at the tensile strut where the LVDTs were installed at the shear zone was 0.065%.

Figure 14. Failure behavior of the first control slab S1.

Two minor cracks were observed during the test in specimen S2 at the roller support location due to not leveling the surface in the ancho at a load equal to 50 kN and 130 kN. At 262 kN, a sudden failure in the shear zone at the hinged support was observed as a shear tension failure. A slippage of the pre-stressing strands occurred at the failure stage. An inclined shear crack was observed at an angle equal to 33 degrees and started at 127 mm from the slab edge. The maximum crack width was 15 mm, as shown in Figure 15. The strain at the compression strut where the LVDTs were installed at the shear zone was −31.6%. However, the strain at the tensile strut where the PI-gauges were installed at the shear zone was 15.9%.

Figure 15. Failure behavior of the second control slab S2.

The observed crack in specimen S3 was a sudden failure in the shear zone at a load of 41.9 kN. This was due to shear tension failure. Slippage of the pre-stressing strands occurred at the failure stage. An inclined shear crack was observed at an angle equal to 39 degrees and started at 180 mm from the slab edge. The maximum crack width was 10 mm, as shown in Figure 16. The strain at the compression strut where the LVDTs were installed at the shear zone was −0.35%. However, the tensile strain, which is a perpendicular strut where the PI gauges were installed at the shear zone, was 13.5%.

Figure 16. Failure behavior of slab S3.

The observed crack in specimen S4 was a sudden failure in the shear zone at load 384 KN. A separation between the slab and the concrete topping was observed. There was a bottom crack beside the hinge support. This was due to shear tension failure. Slippage of the pre-stressing strands occurred at the failure stage. An inclined shear crack was observed at an angle equal to 26 degrees and started at 100 mm from the slab edge. The maximum crack width was 30 mm, as shown in Figure 17. The increase in HCPPRC slab strength was 46%, while the percentage increase in total thickness/original thickness was 28.33%. While the increase in slab strength from the previous study (Girhammar & Pajari, 2008) was 35%, the depth increase was 90/200, or 45%. The strain at the compression strut where the LVDTs were installed at the shear zone was −27.8%. However, the strain at the tensile strut where the PI gauges were installed at the shear zone was 10.21%.

Figure 17. Failure behavior of slab S4.

The first crack in specimen S5 was observed in the shear zone at a load of 350 kN. Dependence between the steel plates and the specimen was observed at the failure stage. There was observed a separation between the slab and the steel plates as a shear tension failure. A slippage of the pre-stressing strands occurred at the failure stage. An inclined shear crack was observed at an angel equal to 42 degrees and started at 120 mm from the slab edge. The maximum crack width was 35 mm, as shown in Figure 18. The compression strain at the compression strut where the LVDTs were installed at the shear zone was 4.43%. However, the tensile strain at the tensile strut where the PI gauges were installed at the shear zone was 1.14%. The failure load recorded 469 kN with an increasing percentage in shear capacity of about 46%, while the Girhammer study (Girhammar & Pajari, 2008) of 80 mm topping concrete thickness recorded 35%. Also, the failure shape of shear tension was identical for the two studies.

Figure 18. Failure behavior of slab S5.

The first and second cracsk in specimen S6 were in the shear zone at the roller support at a load of 300 and 310 kN, respectively. The third crack was in the shear zone, also at a load of 322 kN. This was due to shear tension failure. There was an observed sliding in the reinforcement strand at 377 kN. The cracking angle was 43 degrees and started at 130 mm from the slab edge. The maximum crack width was 25 mm, as shown in Figure 19. The strain at the compression strut where the LVDTs were installed at the shear zone was −0.275%. However, the strain at the tensile strut where the PI gauges were installed at the shear zone was 26.68%.

Figure 19. Failure behavior of slab S6.

The first crack in specimen S7 was in the shear zone at the hinge at a load of 141 kN. The second crack was a horizontal crack above the flange at a load of 290 KN. The third crack was in the shear zone at the hinge at a load of 305 kN. This was due to shear tension failure. Slippage of the pre-stressing strands occurred at the failure stage. The cracking angle was 41 degrees and started at 130 mm from the slab edge. The maximum crack width was 35 mm, as shown in Figure 20. The strain at the compression strut where the LVDTs were installed at the shear zone was −0.612%. However, the strain at the tensile strut where the PI gauges were installed at the shear zone was 23.64%.

Figure 20. Failure behavior of slab S7.

The first crack in specimen S8 was under the secondary beam at a load of 334 kN. The second crack was freckles in the webs at a load of 383 kN. This was due to shear tension failure. The third crack was a sudden failure in the shear zone in addition to longitudinal failure in the slab due to the separation between the specimen and the retrofitting concrete at a load of 423 kN. A slippage of the pre-stressing strands occurred at the failure stage. The cracking angle was at 27 degrees and started at 120 mm from the slab edge. The maximum crack width was 26 mm, as shown in Figure 21. The strain at the compression strut where the LVDTs were installed at the shear zone was −37.4%. However, the strain at the tensile strut where the PI gauges were installed at the shear zone was 0.108%.

Figure 21. Failure behavior of slab S8.

4.2. Load-deflections response

Figure 22 shows the load-deflection curves of the specimens S1, S2, S3, S4, S5, S6, S7, and S8 for the deflection in the middle of the slabs with the maximum measured load (P_max) on the slabs. It could be seen that specimens S1, S4, S5, S6, S7, and S8 had initial cracks before failure, while specimens S2 and S3 had a sudden failure without any initial cracks due to strengthening S2 in flexure only and negative shear strengthening of S3.

Figure 22. The load-deflection curves of (a) S1 and (b) S2, S3, S4, S5, S6, S7 and S8.

4.3. Comparisons between all specimens

The maximum measured loads of the specimens S1, S2, S3, S4, S5, S6, S7, and S8 were 271.4, 262, 41.9, 384, 469, 420, 305, and 423 kN, respectively. The failure load measured for specimen S3 was less than that of specimen S2 by 84%, which was strengthened with screws and nuts. This is mainly due to disbonding between the concrete slab and the screws with nuts. This made each web work alone. The maximum load measured for specimen S4 was higher than that of specimen S2 by 46.6%, which was strengthened with an 85-mm concrete topping. This is mainly due to the good behavior between the retrofitting slab and the concrete topping. The maximum load measured for specimen S5 was higher than that of specimen S2 by 79%, which was strengthened by filling about 50% of voids with concrete. This is mainly due the good behavior between the retrofitting slab and the filling concrete. The maximum load measured for specimen S6 was higher than that of specimen S2 by 60.3%, which was strengthened by filling about 50% of voids with concrete and screws with nuts. This is mainly due to the good behavior between the retrofitting slab and the retrofitting method. The maximum load measured for specimen S7 was higher than that of specimen S2 by 16.4%, which was strengthened with CFRP sheets. This is mainly due to the good behavior between the retrofitting slab and the retrofitting method. The maximum load measured for specimen S8 was higher than that of specimen S2 by 61.45%, which was strengthened by filling 100% of the shear voids with concrete. We achieved this result because the area of concrete was so large and qu (the nominal ultimate shear strength) was greater than qu_max (the maximum shear strength). Finally, it is recommended to use 50% filling of voids as the best technique for shear strengthening of HCPPRC slabs.

5. Conclusions

Based on the previous data and strengthening for eight specimens of HCPPRC slabs of thickness 300 mm using different techniques, the following conclusions can be drawn:

• Using screws with nuts only decreased the slab’s concrete strength. It is not preferred to use this method for strengthening HCPPRC slabs.

• Strengthening with an 85-mm concrete topping increased the shear strength of HCPPRC slabs by about 46%.

• Filling about 50% of the shear voids with concrete increased the shear strength of the HCPPRC slab by about 79%, so it is considered the ideal shear strengthening technique for HCPPRC slabs.

• Filling about 50% of the shear voids with concrete in addition to screws with nuts was an approved method for strengthening HCPPRC slabs because the behavior was closer to being ductile and increased the shear slab concrete strength by about 60.3%.

• Strengthening the HCPPRC slabs with Carbon Fiber Reinforced Polymer (CFRP) sheet as stirrups is a simple technique for strengthening the HCPPRC slabs, and it increased the shear slab concrete strength by about 16.4%.

• Filling all shear voids with concrete increased the shear strength of HCPPRC slab by about 61.45%, but it isn’t an economic technique.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

Thanks to the agency [VEGA 1/0626/22] of the Scientific Grant agency, the Ministry of Education, Science, Research, and Sport of the Slovak Republic and the Slovak Academy of Sciences for the opportunity to publish this post (for APC).

Notes on contributors

Dušan Katunský

Professor Katunský is the author of several book publications, a narrative and original scientific articles published in journals and proceedings. He is an authorized engineer and expert in building constructions and building physics. He is a member of the editorial boards of conferences and journals reviewer of many articles.