1 Introduction

During last decades, different techniques were investigated to increase flexural capacity and shear behaviour of reinforced concrete beams. For example using, high-performance concrete [1], fiber reinforced polymers (FRP) [2], and prestressed concrete [3], etc. Most of these methods have showed significant improvement in test results. These methods mainly improve the behaviour of RC beams by using high strength reinforcement and/or improving the mechanical properties of concrete mix. Traditional steel trusses were widely used when large-spans are required, but it has low durability, poor fire resistance, and needs regular maintenance routines. Recently, researchers focus on hybrid steel trussed reinforced concrete beams (HSTRCB), where a prefabricated steel truss is embedded into concrete beam. These researches tried to overcome the previous mentioned problems [4], by impeding the steel truss in a newly cast concrete beams, to form a composite beam. Composite beam with impeded steel truss technique has been used for many years in Europe as it has numerous advantages. One of the main advantages for this technique is to solve the problem of instability in many structure cases especially in connections [5]. Also, this type of beams have the ability to carry its own weight, so it reduces supports during construction which saves time and effort, and improve the shear and flexural capacity of reinforced concrete beams compared to standard ones [6,7,8].

[8] developed an analytical formula based on the numerical analysis to evaluate the critical moment for lateral torsional buckling of truss members, which may suffer buckling due to loads of the weight of precast slab before hardening of used concrete, so at this stage the truss system acts as a steel structural element without composite behaviour till hardening of concrete.

[7] replaced the traditional reinforcement of beams with steel truss system, where bottom reinforcement acted as bottom chord of truss, and top reinforcement acted as upper chord, while inclined bars acted as web reinforcement (stirrups) which welded at the bottom and upper reinforcing bars as shown in Fig. 1a. The effect of this technique on flexural and shear capacity of the beams through experimental tests were investigated by changing the web reinforcement and depths. The results showed that, beams with new reinforcing technique had wider cracks formed compared to that of traditional reinforced beams due to the longer spaces between web bars and their smooth surfaces.

Fig. 1
figure 1

Specimens details for previous researches of RC beams with impeded truss

Full size image

[9] replaced the traditional reinforcement of beams with steel truss system, where bottom reinforcement was replaced by steel plate, and two steel bars were used as upper reinforcement, inclined bars acted as web reinforcement (stirrups) which welded in the bottom and upper reinforcing bars as shown in Fig. 1b. The shear capacity of this composite beams using experimental and numerical modelling was investigated. Based on the FEA models, a formula to evaluate the shear behaviour of the beam was developed. On the other hand, the effect of changing the bar diameter of the web reinforcement on the crack depths and propagation was investigated.

Detailed investigation for the flexural behaviour of trussed beams was conducted by Colajanni et al. First, the weld connection between truss members was studied by testing truss system not surrounded by concrete under push-out tests, then the truss system which embedded in concrete was experimentally tested to investigate the mechanism of stress transferring between steel truss elements and concrete. Finally, FEA models were developed to derive an analytical designing formulas for these types of composite beams [6, 10, 11].

Nan Zhang et al. added a steel angle truss into the core of traditional reinforced concrete beams with traditional reinforcement bars and investigate the shear performance of the tested beams experimentally as shown in Fig. 1c. The experimental results further demonstrate that embedding the steel trusses in reinforced concrete beams improved shear capacity of RC beams by 80% in case of beams of small span to depth ratio [12].

Kun Wang et al. used a steel truss skeleton as a reinforcement in concrete beams. Bottom and top reinforcement were replaced by hollow square tubes filled by grouted materials, and the web reinforcement by steel rods as shown in Fig. 1d. The effect of steel ratio of upper and lower chords, strength of concrete, and distance between steel tube truss joints on the vertical load–deflection curves were investigated. The results indicated that the steel tubes of the upper chords in the composite beams did not reach yield for full sections due to the structural requirements while, the lower chords reached the steel yield strength first then the upper concrete crushed subsequently, showing a mode of failure similar to that of the under reinforced concrete beam. The steel ratio of lower chords mainly affected the flexural capacity of the normal section among the investigated factors [4].

Salah et al. investigated the effect of changing of web reinforcement configurations on the flexural behaviour of RC beams with impeded truss. Crack pattern and the propagation of cracks through different loading periods were observed. It was concluded that the beams with N-shape truss gave the best flexural behaviour compared to traditional reinforced beams [13].

The above mentioned researches concluded that trussed beams technique improves shear and flexural behaviours of reinforced concrete beams. Most of the previous studies were concerned on changing the steel configurations as shown in Fig. 1.

The current research paper experimentally investigates a new steel reinforcement configuration for reinforced concrete beams. The bottom and top reinforcement were replaced by steel angles, while web reinforcement (stirrups) were replaced by steel batten plates. The flexural behaviour, ultimate load capacity and crack pattern were investigated for the tested beams using two points flexural loading test.

2 Methodology

The aim of the current study is to determine the effect of using smooth steel truss instead of regular steel bars (HSTRCB) in terms of the flexural capacity of beams. Different smooth truss typology were used as reinforced rebar replacement in the RC beams to determine the significant truss configuration that has highest flexural capacity. Accordingly, a doubly reinforced beams with impeded steel truss was designed having a compression and tension reinforcement ratio of 0.64%. The top steel angles with cross section of 40 × 40 × 4 mm were used, while the dimensions of bottom steel angles section is 50 × 50 × 5 mm. All specimens were designed and fabricated to resist bending moment, avoiding shear collapse during the entire stages of experimental loading.

The specimens were designed to have the same ultimate moment capacity, using first principle equations, according to the ACI 318 [14]. Taking into consideration that the centroid of the angles and the traditional steel bars are not the same, and the yield strength of steel material used for the rebars and steel angles are not equal as shown in Table 2.

Designing procedure for the double reinforced section (Fig. 2) is occurred by enforcing the equilibrium as following:

Fig. 2
figure 2

Section equilibrium of doubly reinforced concrete beam

Full size image
$${A}_{s}{f}_{s}=\left(0.85\right){f}_{c}{\prime}b{\beta }_{1}c + {A}_{s}{\prime}({f}_{s}{\prime}-0.85 {f}_{c}{\prime})$$
(1)
$${A}_{s}{f}_{s}=\left(0.85\right){f}_{c}{\prime}b{\beta }_{1}c + {A}_{s}{\prime}({E}_{s}{\varepsilon }_{cu}\left(\frac{c-{d}{\prime}}{c}\right)-0.85 {f}_{c}{\prime})$$
(2)

The value of c = 63.63 mm determined using Eq. 2.

Hence, the ultimate moment was calculated using Eq. 3

$${M}_{u}= {A}_{s}{f}_{s} \left( d-\frac{{\beta }_{1}c}{2} \right)- {A}_{s}{\prime}({E}_{s}{\varepsilon }_{cu}\left(\frac{c-{d}{\prime}}{c}\right) ({d}{\prime}+e-\frac{{\beta }_{1}c}{2})$$
(3)

Now the ultimate moment for the HSTRCB was determined (100.1 kN m) using the first principle and the steel rebar for the reference sample could be calculated using the same procedure. Steel rebar area and the value of c are the two unknowns in the reference sample. They could be determined using first principle equations.

3 Experimental work

Five reinforced concrete beams with the same concrete cross-sections and span were casted and tested under two point bending load test. Same concrete mix was used to cast all the R.C. beams. The concrete mix proportion, concrete mechanical properties, steel mechanical properties, specimen details, and test procedure are presented in the following sections.

3.1 Concrete mix proportion and mechanical properties of used material

Table 1 illustrates the concrete mix proportion. The compressive strength was measured by taking the average of three tested cubes of concrete after 28 days of curing. Table 2 shows the mean experimental values of the mechanical properties of concrete and steel. The yield strength of steel was tested in lab using a calibrated universe tensile testing machine.

Table 1 Concrete Mix Proportion
Full size table
Table 2 Material Properties of Concrete and Steel Parts
Full size table

3.2 Specimens details

Flexural behaviour of hybrid steel-truss reinforced concrete beams (HSTRCB) was studied experimentally. Concrete beams having the same cross-section (220 mm width, 350 mm depth) and span length (3400 mm) were casted and tested. The specimens were classified into two groups according to longitudinal and transversal reinforcement details. The first group, named as group I, is the RC standard beam with traditional rebar reinforcement (SB1), while the second group, named as group II, is the HSTRCB with impeded steel truss (NB5, WB5,V29B5, V20B5). Reinforcement details for specimens are shown in Table 3. Figures 3a illustrates gathering and welding of steel skeleton, while Fig. 3b shows the specimens after casting.

Table 3 Specimens’ Details
Full size table
Fig. 3
figure 3

Preparing of specimens

Full size image

Specimen SB1, shown in Fig. 4, was designed as a reference beam to be used in comparing the behaviour of the HSTRCB. All specimens have reinforcement ratio less than the maximum ratio of the double reinforced sections according to ACI 318 [14] to enforce ductile failure. Steel arrangement of batten plates with fixed cross section (plates 30 × 4 mm) that act as web reinforcement was varied according to spacing, shape, and inclination angle. Beams component were designed, drawn, assembled and welded as shown in Figs. 4, 5, 6, 7 and 8.

Fig. 4
figure 4

Beam SB1 details [16]

Full size image
Fig. 5
figure 5

Beam NB5 details

Full size image
Fig. 6
figure 6

Beam WB5 details

Full size image
Fig. 7
figure 7

Beam V29B5 details

Full size image
Fig. 8
figure 8

Beam V20B5 details [16]

Full size image

The stress transfer between concrete block and imbedded reinforcement depends mainly on the bond strength between concrete and steel. Conventional specimen (SB1) should exhibit higher bond strength due to the deformed outer surface of the reinforcement bars. On the other hand, HSTRCB specimens experience lower bond strength compared with SB1 due to the smooth surface of the used steel angles, which installed as a longitudinal lower and top chord of the imbedded truss, as shown in Figs. 5, 6, 7, 8. Steel truss typology, in HSTRCB, with diagonal and vertical batten plates were expected to participate in the transferring action between steel truss and concrete block. That causes a balance of bond strength for these specimens compared with conventional reinforcement (SB1). The aforementioned criteria is one of the key points that will be highlighted in this research. Moreover, bond strength between steel truss and concrete block needs to be improved by means of shear studs in HSTRCB specimens to enhance the stress transfer between concrete and the imbedded steel truss.

3.3 Test setup

Two point load test was used to examine the flexural capacity of the beams, to enforce constant moment at the mid span. The tested beams are simply supported at its ends. The first end is hinged support and the other is roller support with clear distance of 3400 mm. The specimens were loaded, using a calibrated load cell of 2000 kN load capacity, at two loading points with 1000 mm spacing apart as illustrated in Fig. 9. The vertical displacement of the tested specimens were measured during test process using three linear variable transducer (LVDT). One was located at the extreme tension fiber of the mid span of the tested specimens, while the other two LVDTs were installed at the extreme tension fiber of the beam, directly under the two loading points, as shown in Fig. 9. Three strain gauges were installed on the longitudinal top and bottom reinforcement, the first two were installed on the main steel bottom chord at mid-span, while the third one was installed at the top steel chord at mid span as shown in Figs. 4, 5, 6, 7 and 8. The measured data were recorded by a data logger connected with computer system program using “lab view” software. Figure 9 shows a sketch of test setup and instrumentation of all tested specimens.

Fig. 9
figure 9

Test setup

Full size image

4 Results and discussions

4.1 Load–displacement curve

According to calculations all the beams were designed to have same theoretical ultimate moment of 100.1 kN m using Eq. (3), which requires a corresponding load of 166.9 kN (theoretical ultimate load). Load–deflection curves and modes of failure of all the specimens were shown and illustrated in Figs. 10 and 11, respectively. By summarizing the curve results, it was shown that, specimen SB1 has the highest ultimate load of 188 kN with corresponding displacement of 134 mm, while, specimen NB5 has ultimate load of 186 kN, which nearly equal to that of specimen SB1, and corresponding displacement of 135.2. NB5 also experienced slipping between the concrete cover and steel angles at displacement of 150 mm (experimental observation). Specimen WB5 has ultimate load smaller than specimen SB1 by about 5.3% with ultimate load of 178 kN. Specimens V20B5 and V29B5, showed same behaviour and ultimate load values, where the ultimate load values for the two specimens were 172 kN, nearly above the theoretical ultimate load (which equal to 166.9kN) and decreasing of 8.5% compared to specimen SB1. It could be concluded that the variation of the horizontal distance between vertical plates does not have significant impact on flexural behaviour of the tested beams. These specimens failure pattern was identical to those of the WB5 and NB5 as shown in Fig. 11.

Fig. 10
figure 10

Load–displacement curve of all specimens

Full size image
Fig. 11
figure 11

Failure of the specimens (Front view)

Full size image

All major cracks could be classified as flexural tension cracks. Moreover, specimen SB1 showed more crack number and subsequently less crack widening compared to HSTRCB. Hence, crack width for HSTRCB is greater than of traditional beam SB1 at the same loading value. This could be concluded due to the week bond between steel truss and concrete core.

It was noticed that HSTRCB specimens exhibited lower Steel yielding point than SB1, which displayed higher steel yielding point followed by local rebar buckling at the top reinforcement and spalling of concrete in the compression zone, forming a plastic hinge, see Fig. 12b. Plastic hinges are regions where beam section can rotate without increasing of bending moment resistance and subsequently the resistance to the applied load decreases, thus dissipating energy during loading. Unrestricted plastic flow occurs at the mid-span section where large increases in deformation take place without increasing in load. The tested beam therefore behaves as two rigid beams connected by a plastic hinge which allows them to rotate relative to each other [1]. Specimen SB1 only shows decreasing in load resistance after formation of the plastic hinge up to failure.

Fig. 12
figure 12

Failure of the top of tested specimens

Full size image

It was recorded during loading process of all HSTRCB specimens that, concrete at compression side surrounding steel angle at top chord was spalled at early loading stage (this issue resulted from the weak bond between steel angles and concrete) compared with specimens SB1,. On the other hand, no plastic hinges occurred in HSTRCB specimens, while no load decaying was observed until the end of test. Huge mid -span cracks (more than 10 mm wide) accompanied with large displacement and rotation angle at the supporting ends were observed during testing HSTRCB. This indicates that HSTRCB specimens exhibited better ductility after steel yielding compared to the standard beam. The recommended reason for this observation is that impeded steel truss in HSTRCB specimens started to resist the applied load individually. Catenary action has occurred in HSTRCB after surrounding concrete was completely failed. This action was not observed in the specimens with conventional reinforcement.

4.2 Cracking moment

Analytical deflection evaluation mainly depends on cracking moment of the specimens. Gross moment of inertia (Ig) could be used for deflection calculation below the cracking moment, ignoring the existence of reinforcement. Cracking moment is determined using the following equation:

$${M}_{cr}=\frac{{f}_{r}.{I}_{g}}{{y}_{t}}$$
(4)

where: \({f}_{r}\) is the modulus of rupture, \({y}_{t}\) is the depth from the extreme tensile fiber to the neutral axis and \({I}_{g}\) is the gross second moment of area.

First crack load was determined by the change in the slope of the Load–Deflection curve and assured by visual monitoring during testing.

Since, there is no difference between all beams cross section, the gross second moment of area should have the same value, ignoring reinforcement. That means all specimens should have same first cracking moment. That did not occur during testing as shown in Fig. 13. As a result, transformed un-cracked moment of inertia was used to calculate the theoretical cracking moment of the tested specimens.

Fig. 13
figure 13

Experimental first crack load of the tested specimens

Full size image

A comparison between un-cracked transformed sections and gross section is required to verify the experimental cracking moment results. Table 4 showed the moments of inertia of tested specimens (SB1 and HSTRCB), corresponding cracking moment and cracking load, calculated using the properties of section compared with the experimental cracking loads.

Table 4 The moments of inertia of the tested specimens
Full size table

Cracking load could be theoretically calculated using either transformed un-cracked section or gross section. If the cracking load for specimens is compared with experimental cracking load, it is clearly obvious that gross moment of inertia could be used to determine the first crack load for traditional reinforced concrete beam SB1. Contrarily, transformed moment of inertia for un-cracked section could be used to evaluate the cracking load for HSTRCB specimens. Accordingly, it increases the cracking moment and first crack load as shown Table 4 due to the participation of the impeded truss, so transformed moment of inertia is essential to be used in cracking moment calculations instead of gross moment of inertia in HSTRCB specimens.

4.3 Cracked moment of inertia and initial stiffness

After occurring of the first crack during loading of specimen, deflection is controlled by beam stiffness, which mainly depends on the cracked moment of inertia. The cracked moment of inertia in reinforced concrete beams is member cracking consequent when neutral axis varies between cracks along the beam span. A degradation in beam stiffness happens as the cracking propagate. Hence, experimental moment of inertia during the period of testing was evaluated using elastic deformation theory, Eq. 5.

$${I}_{{\text{cr}} {\text{exp}}}= \frac{{P}_{y}.a (3{l}^{2}-4{a}^{2})}{48{E}_{c }{\Delta }_{{\text{exp}}}}$$
(5)

where P is the applied load, α is the shear arm, l is beam clear span and Δ exp is the mid span measured deflection.

As shown in Fig. 14a, b, HSTRCB improved stiffness of the cracking section. The steel angles in HSTRCB acquired the cracked beam cross section higher inertia than specimens with conventional reinforcement. Which in return significantly improves cracked moment of inertia compared to the ordinary reinforced beam (SB1). Figure 15 showed the experimental normalized cracked moment of inertia calculated along the period of the test. It was found that the experimental cracked moment of inertia for the HSTRCB increased significantly at the designing moment (0.85 Mu).

Fig. 14
figure 14

The initial stiffness of the tested specimens

Full size image
Fig. 15
figure 15

The experimental moment of inertia

Full size image

4.4 Post cracking stiffness

The embedded steel truss skeleton provide additional flexural stiffness for the reinforced concrete beam and enhanced the tensile and compressive strength of critical failure section of reinforced concrete beams, which makes the deformation of the specimens showing ductile behaviour [12].

Figure 14a illustrates initial and Post cracking stiffness regions. The Post cracking stiffness can be represented as the slope of the line in the elastic zone after cracking to the yielding point as illustrated in Eq. 6.

$${\text{Post}} {\text{cracking}} {\text{stiffness}} = \frac{{P_{{\text{y}}} - P_{{{\text{cr}}}} }}{{\Delta_{{\text{y}}} - \Delta_{{{\text{cr}}}} }}$$
(6)

where Py and Pcr are the yield load, and cracking load, respectively. While, \({\Delta }_{y}\) and \({\Delta }_{cr}\) are the associated displacement corresponding to yield and cracking loads, respectively.

As shown in Fig. 16 the Post cracking stiffness of all the HSTRCB was improved compared to the standard reinforced beams. Post cracking stiffness showed that specimen NB5 enhanced the post cracking stiffness by 77% than SB1.

Fig. 16
figure 16

Post cracking stiffness of The Tested Specimens

Full size image

4.5 Energy absorption

The area under the load–displacement curve represents the energy absorption capacity of the tested specimen. The absorbed energy capacity, was calculated up to deflection of 200 mm for all specimens.

The results showed that HSTRCB improved energy absorption of RC beams, where the energy absorption of WB5 and NB5 were greater than that of the standard reinforced concrete beam by 18.8% and 23.6%, respectively as shown in Fig. 17. The main reason of that is once the reinforcing steel skeleton started to fail, the membrane action of the steel truss members took over completely, as a result a noticeable enhancement in trussed beams with inclined web reinforcement energy absorption values occurred.

Fig. 17
figure 17

The energy absorption of the tested specimens

Full size image

5 Conclusions

This research investigated the flexural behaviour of hybrid steel-truss reinforced concrete beams (HSTRCB). Five reinforced concrete beams with the same cross-section and span were casted and tested under two point load test. First beam is ordinary reinforced concrete beam while the other four HSTRCB were designed having identical steel angles, as compression and tension reinforcement, with different shear reinforcement alignment, and spacing. It should be emphasized that the following conclusions are based on experimental data:

  1. 1.

    The reinforcing technique used in HSTRCB significantly improved the cracking moment and cracked moment of inertia compared to standard concrete beam, specially N-shaped truss configuration (NB5).

  2. 2.

    Since de-bonding problems were found in all of the tested beams with impeded steel truss, where, the concrete at compression side surrounding the top chord steel angle was spalling at early loading stage compared with specimens SB1 (specimens with conventional steel reinforcement). It is important that the smooth surface of steel angles be taken into account while constructing this kind of RC beams to increase the bonding between the impeded steel trusses and surrounding concrete. This bonding should be enhanced by adding dowels along of the surface of the steel angels or used other techniques aims to same precautions during construction. That requires more future studies.

  3. 3.

    All beams exhibited linear behavior until the reinforcing steel skeleton began to yield, after that point they entered a nonlinear (inelastic) zone. Standard beam, exhibited load decay after reaching ultimate load. On the other hand, HSTRCB entered a new linear zone in which the membrane action of steel truss members carried the load entirely. The recommended reason for this observation is that impeded steel truss in HSTRCB specimens started to resist the applied load individually. Catenary action has occurred in HSTRCB after surrounding concrete was completely failed. This action was not observed in the specimens with conventional reinforcement.

  4. 4.

    HSTRCB specimens exhibited better ductility after steel yielding compared to the standard beam with no softening behavior happened in the load and specimens were able to pursue load resistance.