Effect of anchorage number and CFRP strips length on behavior of strengthened glulam timber beam for flexural loading

Abstract

Laminated wooden beams are more preferred in the production of wooden structures than solid timber beams because they have a higher load-carrying capacity and allow larger openings to be used in the structure. The widespread use of wooden structures and the increasing size of the structures have revealed the need for strengthened laminated wooden beams and increase their ultimate load capacity. It has become necessary to develop reinforcement details to increase the ultimate load capacity of laminated wooden beams in wooden railroads or highway bridge beams, where the traffic load increases, especially in large wooden structures, in cases where large openings must be passed. Within the horizon of the study, the behavior and performance of three-layer glulam wooden beams strengthened with anchorage and non-anchorage CFRP strips with different bonding length under flexural loading were investigated experimentally. The three-point bending test was applied to glulam timber beam test specimens produced by laminating yellow pine wood material using the polyurethane adhesive. General load-displacement behaviors, ultimate load capacity, initial stiffness, displacement ductility ratios, and energy dissipation capacities were obtained. The increase in the bonding length of the CFRP strips used for strengthening in the glulam timber beam specimens and the use of CFRP fan type anchors at the strip ends increased the ultimate load capacity and initial stiffness values of the wooden beams, as well as the displacement ductility ratios and energy dissipation capacity values.

Keywords

CFRP CFRP fan type anchor flexural capacity glulam timber beam strengthening

Introduction

Wood is an important building material that can be obtained from nature, controlled sustainably with industrially grown raw material, and also has very high strength compared to its weight. In the last 20 years, there have been major developments in constructing buildings using wood entirely. Wood is not only used as a massif wood in buildings, but many different types of engineered wood composites have been developed and used in the manufacture of wooden structures. In particular, glued laminated timber (glulam) is widely preferred among engineered wood composites. The development and widespread preference of Glulam production technique as the beams and columns that form a carrying system in wooden structures have enabled the development of wooden structures with a larger and higher load-carrying capacity. The developing and widespread use of wooden structures have allowed it to be used in different structures such as industrial structures, sports facilities, road or railway bridges, storage tanks, transmission towers, and larger structural systems.

Analyzing the literature is revealed that reinforcing beams produced with glulam production techniques are strengthened with different techniques. Studies in the literature have been revealed that elements made of materials such as steel and aluminum, steel plate, or high-strength cable elements with pre-tensioning are used for reinforcement (Haller et al., 2015; Jasienko and Novak, 2014; Luca and Marano, 2011; Mark, 1961; McConnell et al., 2014; Triantafillou and Deskovic, 1992). Later, as composite materials became more and more widespread in reinforcing of reinforced concrete or masonry structures, it was observed that fiber-reinforced polymer (FRP) composites were also used for strengthening wooden structures (Ambrisi et al., 2014; Andor et al., 2015; Andre, 2006; Borri et al., 2013; Cabrero et al., 2010; Conde et al., 2015; Corradi et al., 2015; Fossetti et al., 2015; Gentile et al., 2002; Gilfillan et al., 2001, 2003; Kasal et al., 2004; Khelifa and Celzard, 2015; Khelifa et al., 2015; Li et al., 2014; Lu et al., 2015; Negrao et al., 2011; Raftery and Kelly, 2015; Schober et al, 2015; Sena-Cruz et al., 2013; Yeboah et al., 2013).

Much research has been accomplished in improving the strength and stiffness of wooden beams using FRP composite (Franke et al., 2015). Gezer and Aydemir (2010) determined the increase in compressive and flexural strength of fir and pine wood reinforced with carbon fiber reinforced polymer (CFRP). The bending behavior of wooden beams reinforced with hybrid FRP (HFRP) has been shown by Yang et al. (2013). Li et al. (2009) investigated the bending performance of wooden beams retrofitted with CFRP composite sheets and observed the increase in bending strength with the number of composite sheet layers. Borri et al. (2005) increased the flexural stiffness and strength of existing wooden elements by reinforcing FRP materials. To estimate the behavior of timber beams strengthened with CFRP composites Kim and Harries (2010) presented a FEM approach and compared that with the experimental results. Alhayek and Svecova (2012) found that the stiffness increase is minimal with the FRP reinforcement of wooden beams and largely depends on the span-depth ratio of the beams. Naghipour et al. (2011) found that wood-plastic composite (WPC) beams retrofitted with CFRP and GFRP strips on the tension surface positively affected the strengthening and restoration of WPC beams. A simple and efficient finite element design that can predict the load-carrying capacity of wooden beams reinforced against bending loading with FRP strips and FRP bars has been developed by Valipour and Crews (2011). Hernandez et al. (1997) presented a study on low cost and mass production to produce glulam-FRP beams with high distinguished stiffness and strength in commercial facilities. The need for Carbon fiber reinforced glued laminated beams with high performance and fire resistance caused the development of a new phenolic resin and CF/P composite sheet (pulp paper impregnated) by Ogawa (2000) and suggested the developed CFR-glulams a structural material in wooden structures. Issa and Kmeid (2005) compared the flexural properties of the reinforced and non-reinforced laminated beam. By using pultruded rectangular carbon fiber rods, Johnsson et al. (2006) conducted a probability study on the strengthening of glulam beams and anchorage distance for the system. Raftery and Harte (2011) and Raftery and Harte (2013) studied the reinforcement of low-grade glued laminated timber under the effect of flexure loading reinforced with a pultruded glass fiber polymer composite plate. Riberio et al. (2009) conducted a study by strengthening glulam maritime pinewoods using glass fiber and found useful results. Fiorelli and Dias (2011) conducted an experimental study on FRP reinforced glulam beams and evaluated the results theoretically.

In the review, it is observed that studies on strengthening glulam wooden beams by carbon-reinforced fiber fabrics (CFRP) are included in the literature. However, the reinforcement detail applied to the beams increases the maximum load-carrying capacity and initial stiffness values. It ensures that displacement ductility ratio and energy dissipation capacities are not negatively affected as much as possible. In the strengthening method, it is not enough to increase the load-carrying capacity and stiffness of glulam beams, which is an important part of the carrying system of wooden structures. While these features are improved, at the same time, excessive friability of the beams and the failure mechanisms of the beams should be prevented from being too brittle. The strengthening details applied to the reinforced glulam wooden beams are also very important to create strengthening solutions that will improve displacement ductility ratio and energy dissipation capacities. Within the horizon of this study, an experimental study is planned to create a new strengthening detail that can contribute to the literature for this subject. Within the scope of the experimental study, CFRP strip length change, which is one of the most effective variables to improve the overall load-displacement behavior of glulam wooden beams, has been investigated by maintaining the balance between strength and ductility. Besides, within the intention of this study, the effects of anchorage application, which delay debonding of CFRP strips from the surface in the endpoints where the stress concentrations of CFRP strips intensify, is another innovative study subject that will contribute to the literature which is planned to be examined.

In this study, 10 test specimens, one of which is a non-reinforced reference glulam wood beam, were produced by using yellow pine wood material. Glulam wooden beams were tested under the influence of three-point flexure loading and general load-displacement behaviors, maximum load-carrying capacity, initial stiffness values, displacement ductility ratios, and energy consumption capacities were obtained. The results of the wooden glulam beams strengthened with CFRP strips are compared without strengthened reference test specimen. Besides, finite element models were created using ABAQUS finite element software for strengthened, laminated wooden beams, which experiments were carried out in the study. The results obtained from the experimental results and numerical finite element models were compared and interpreted to what extent successful results were obtained by computer simulation.

Experimental study

Test specimens and materials

As part of the experimental study, glulam wood beam test specimens were produced using yellow pine (Pinus sylvestris) wood material. Engineered wood composites test specimens produced with the glulam technique in the experimental study were then strengthened against flexure loading by using carbon-reinforced fiber fabric (CFRP) strips. In the experimental program, one non-reinforced and nine reinforced, 10 test specimens were produced. These specimens were tested under the influence of a three-point flexure loading, which was increased until collapse happen, monotonically. The variables examined in the experimental study are the length of the CFRP strip and the number of anchorages used at the endpoints of the CFRP strips, which are glued to the tensile surface of glulam wooden beams for reinforcement. In the study, for reinforcement, three different lengths of CFRP strips; 1500, 1200, and 900 mm, adhered to the tension surface of the specimens. It is intended to use CFRP fan type anchorages (produced by CFRP strips), which are widely used to strengthened reinforced concrete and masonry structures at the endpoints of the CFRP strips. In the literature review, no study was found in which CFRP fan type anchorages were used in the details developed for strengthening of glulam wooden beams. Within the intent of the study, three different types of applications were made in the endpoint of the CFRP strips; non-anchorage, one anchorage, and two anchorages. The properties of the test specimens are given in Table 1.

Table 1.

Properties of test specimens.

Specimen no.	Number of glulam layer	Length ofCFRPstrip (mm)	Anchorageusage ofCFRP stripedge
D-1	3	–	Without anchor
D-2		1500
D-3		1200
D-4		900
D-5		1500	One anchor
D-6		1200
D-7		900
D-8		1500	Two anchor
D-9		1200
D-10		900

Wooden beams are 90 × 90 mm in size and 1710 mm in length. Tests and properties of the test specimens were determined according to EN 13183-1. The geometric dimensions of with anchorage and without anchorage test specimens are given in Figures 1 and 2, respectively. Laminated wooden beams are produced from three layers of 30 mm thickness. During the experiment, loading was applied to the bonding surfaces in the normal direction. As part of the experimental study, Pinus sylvestris timber, widely used in the construction industry, was used. In the timbers used in the tests, attention has been taken to avoid wood defects such as knots, cracks, and fiber defects. The physical and mechanical properties of the wood are shown in Table 2. In the lamination production, the polyurethane adhesive is used. The polyurethane adhesive is highly elastic and has a high resistance to water, oils, chemical materials, and microorganisms. The material to be glued with polyurethane adhesive should be dry and free from dust and oil. The material moisture should be between 8% and 12%. The properties of the polyurethane adhesive used in the study are given in Table 3.

Figure 1.

Test specimens without anchorage (D2, D3, and D4).

Figure 2.

Test specimens with one and two anchorage (D5, D6, D7, D8, D9, and D10).

Table 2.

Mechanical properties of Pinus Sylvestris (yellow pine) wood.

Remarks	Symbol	Value	Unit
Density	ρ_k	490	kg/m³
Density	ρ₁₂	520	kg/m³
Contractioncoefficient	β_r	4.0	%
	β_t	7.7	%
	β_v	12.1	%
Elastic modulus	E_0,mean	11,700	MPa
Shear modulus	G_mean	731.25	MPa
Bending strength	f_m,k	98	MPa
Tension strength	f_t,0,k	102	MPa
Compressionstrength	f_c,0,k	54	MPa

Table 3.

Mechanical properties of the CFRP, epoxy resin, and polyurethane glue.

CFRP properties
Remarks	Value
Weight	220 g/m²
Thickness	0.12 mm
Tension strength	4100 MPa
Modulus of elasticity	231 GPa
Ultimate strain	1.7%
Epoxy resin
Remarks	Value
Density	1.31 kg/L
Mix ratio	White/grey compound = 4/1
Application temperature	Min +10°C, max +35°C
Tension strength	30 MPa
Bending modulus of elasticity	3800 MPa
Polyurethane glue
Compressive strength	60 MPa
Tensile strength	26.2 MPa
Elastic modulus	830 MPa
Elongation at break	13%

In the production of solid one-piece beam samples, the wood was first cut to 105 × 105 × 1800 mm and then kept until 12% humidity at 20 ± 2°C temperature and 65 ± 5% relative humidity without exposure to direct sunlight. The timbers which reached the air-dry humidity were cut in the net size of 90 × 90 × 1710 mm. The laminates that reached equilibrium humidity, a face, and oriels were firstly opened on the planer and then brought into a thickness of 30 mm in the thickness machine. Three 30 mm thick wooden layers are fixed to each other with polyurethane adhesive. After completing the production of Glulam wood beams, CFRP strips adhere to the tensile surfaces of the wooden beams as provided in Figures 1 and 2. In the experimental study, the mechanical properties of CFRP strips and epoxy adhesive products by the Sika Company are given in Table 3. The fan anchorages used in the test specimens are produced by wrapping 80 mm wide 125 mm long CFRP strip piece to 80 mm long 10 mm diameter reinforcement. In the CFRP strips, the 45 mm portion outside the reinforcement is divided into eight pieces of 10 mm width, and the CFRP strip is separated as a fan. In this research, the details of the used anchorages are given in Figure 3. By determining the points where anchorage will be applied in the test specimens, a hole with a diameter of 12 mm and a depth of 80 mm is opened for anchorage application.

Figure 3.

Properties of CFRP fan type anchorage.

Test setup and instrumentations

In the experimental study, the MTS-322 brand with a servo-hydraulic loading system, which can perform load and displacement-controlled tests with a capacity of 500 kN was used. The load was applied to the test specimens with a hydraulic actuator system with a capacity of 500 kN, and the displacement was measured from the midpoint of wooden beam test specimens. The three-point static flexure loading, which is increased, was applied to the wooden beam test specimens until it collapsed, monotonically. The applied loading speed for all test specimens was constant as 0.5 mm/s. During the experiments, tests were carried out by following the load-displacement graph.

Experimental results

The ultimate load capacity, initial stiffness, displacement ductility ratios, and energy dissipation capacity values were interpreted from the load-midpoint displacement graphs of the experimental specimens. Load-midpoint displacement graphs obtained as a result of the experimental study are given in Figure 4. Besides, the results obtained from the experiments and the calculated values are summarized in Table 4. Photographs showing the failure mechanisms and damage distribution of the test specimens are given in Figure 5. The initial stiffness values of the test specimens were calculated at the load level of 10 kN where no slope change occurred in the load-midpoint displacement graphs. Displacement ductility ratio values calculated for experimental specimens are calculated by proportioning the displacement value at the failure point to the displacement value at the ultimate load capacity point. The point of failure for the experimental specimens has been determined as the ultimate load capacity value decreases by 15%, and the ultimate load capacity drops to 85%. The energy dissipation capacity values for the experimental specimens were obtained by calculating the area under the load-midpoint displacement graphs. In calculating the energy dissipation capacities of the experimental specimens, the area under the graphs was calculated by taking the region up to the point of failure. The failure point is the same as used for calculating the displacement ductility ratio.

Figure 4.

Load-midpoint displacement graphs of specimens.

Table 4.

Experimental results.

Specimen no.	Ultimate loadcapacity (kN)	Displacement atultimateload (mm)	Displacement atfailureload (mm)	Initial stiffness(kN/mm)	Displacementductility ratio	Energydissipationcapacity(kN mm)	Failuremode
1	20.53	48.46	48.46	0.60	1.00	620.68	Flexural failure
2	31.58	46.58	46.58	0.89	1.00	869.12	CFRP strip debonding and flexural failure
3	25.60	37.58	46.92	0.82	1.25	832.32	CFRP strip debonding and flexural failure
4	22.81	31.67	42.28	0.69	1.34	643.65	CFRP strip debonding and flexural failure
5	38.83	43.08	52.71	0.92	1.22	1339.34	CFRP strip rupture and flexural failure
6	36.72	42.45	42.45	0.84	1.00	849.97	CFRP strip rupture and flexural failure
7	33.55	37.46	37.46	0.81	1.00	835.15	CFRP strip rupture and flexural failure
8	52.19	41.92	41.92	1.51	1.00	1447.69	CFRP strip rupture and flexural failure
9	42.04	39.90	44.55	1.29	1.12	1267.31	CFRP strip rupture and flexural failure
10	40.49	38.45	38.45	0.94	1.00	861.75	CFRP strip rupture and flexural failure

Figure 5.

Comparison of experimental failure modes and numerical fracture distribution of specimens.

Strengthening type effect

The most important variable examined within the scope of the experimental study is the strengthening method applied to glulam wooden beams with CFRP strips. When comparing the without strengthening reference Specimen-1 with other specimens strengthened with CFRP strips. It has been observed that the applied strengthening technique has greatly influenced the overall load-displacement behavior of glulam wooden beams and improved its performance under the effect of flexure loading. The strengthening method applied with CFRP strips increased the initial stiffness, displacement ductility ratios, energy dissipation capacities, and ultimate load capacity values of glulam wooden beams, although decreased the displacement values at the ultimate load capacity. The ultimate load capacity, initial stiffness, displacement ductility ratios, and energy dissipation capacity values of the test specimens strengthened without anchorage CFRP strips were calculated at 30%, 33%, 19%, and 26%, respectively, larger than the reference test Specimen-1 without strengthened in average. The ultimate load capacity, initial stiffness, displacement ductility ratios, and energy dissipation capacity values of the specimens with one CFRP fan type anchorage at the endpoints of the strip were calculated by an average of 77%, 43%, 7%, and 62% larger than the reference test Specimen-1 without strengthened, respectively. The ultimate load capacity, initial stiffness, displacement ductility ratios, and energy dissipation capacity values of the specimens with two CFRP fan type anchorage at the endpoints of the strip were calculated by an average of 119%, 108%, 4%, and 92% larger than the reference test Specimen-1 without strengthened, respectively. In the strengthening technique, when the results are examined, it has been shown that using anchorages at the endpoints of the CFRP strips and increasing the number of anchorages increase the ultimate load capacity, initial stiffness, and energy dissipation capacities of glulam wooden beams. Also, it has shown that these ratios increase is much more than the test specimens without anchorage. However, in the strengthening technique, anchorages negatively affected the displacement ductility ratios and decreased the rate of increase compared to the reference test Specimen-1 without strengthening. It has been observed that the use of anchors at the ends of the CFRP strips is extremely effective in improving the load-displacement behavior of glulam wood beams in general. Anchorage increases the ultimate load capacity, initial stiffness, displacement ductility ratio, and energy dissipation capacity values all at once. In the strengthening method, the use of CFRP fan type anchors in CFRP strips delayed the debonding of the CFRP strips adhered to the tensile surface for strengthening the glulam wooden beams. The crushing occurs in the compression zone; the glulam beams have the ultimate load capacity, initial stiffness, and energy dissipation capacity values increased significantly. CFRP fan type anchors are placed in the strip end regions where the stress concentrations occur in the CFRP strips and the maximum moment zone. This application prevents the CFRP strips from peeling off the tensile surface, glulam wooden beams, and good performance has been achieved in the applied strengthening method.

CFRP fan type anchorage effect

Other variables examined within the scope of this experimental study is the number of CFRP fan type anchorages used at the endpoints of the CFRP strips adhered to the tensile (lower face) surface of glulam wooden beams for strengthening. In the experimental program on the wooden glulam beams, a total of three different types of strengthening were applied, one and two anchorages at the ends of the CFRP strips and without anchorage test specimen. The ultimate load capacity, initial stiffness, and energy dissipation capacity values of the test specimens with one anchorage at the endpoint of the CFRP strips were obtained by an average of 38%, 8%, and 29% greater than the test specimen strengthened with just CFRP strip, respectively. The displacement ductility ratios of the test specimens with only CFRP strip for strengthening on the tensile surface of the glulam wooden beams were calculated by an average of 12% larger than the test specimens with one CFRP fan-type anchorage at the endpoints of the CFRP strips. Using anchorages at the endpoints of the CFRP strips has a positive effect on the ultimate load capacity, initial stiffness, and energy dissipation capacities, while displacement ductility ratios have reduced. The ultimate load capacity, initial stiffness, and energy dissipation capacity values of the test specimens with two anchorages at the endpoint of the CFRP strips were obtained by an average of 69%, 54%, and 51% greater than the test specimen strengthened with just CFRP strip, respectively. The displacement ductile ratios of the test specimens with only CFRP strip were calculated by an average of 15% larger than the test specimens with two CFRP fan-type anchorages at the endpoints of the CFRP strips. It was seen that the increase in the number of CFRP fan type anchorages at the endpoints of the CFRP strips positively affected and significantly increased the ultimate load capacity, initial stiffness, and energy dissipation capacity values. While negatively decreased the displacement ductility ratios. In the strengthening method developed, the increase in the number of anchors used on CFRP strips had positive effects on the transfer of the axial tensile force generated by the bending effect of the CFRP strip adhered to the tensile surface of glulam wooden beams. The tensile force occurring in the CFRP strips in the test specimens without anchoring was met only by the epoxy used for bonding the strips to wooden beams. However, with the use of CFRP fan type anchors and increasing their number, the shear stress distribution occurring at the interface due to the axial tensile force on the CFRP strips was met by the anchors. The ultimate load capacity, initial stiffness, and energy dissipation capacity of the experimental specimens increased significantly. Increasing the number of anchors carried more shear stress, causing CFRP strips adhered to the tensile surface to carry much more axial tensile force. The overall load-displacement performance of glulam timber beams was significantly improved before the strips peeled off the surface.

CFRP bonding length effect

Another variable examined within the scope of the experimental study is the length of the CFRP strips adhered to the tension surface of glulam wooden beams for strengthening. Within the extent of this research, three different lengths of CFRP strips; 1500, 1200, and 900 mm were glued to the tension surface (bottom face) of glulam wooden beams. As the lengths of the CFRP strips were used for strengthening increase, the ultimate load capacity, initial stiffness, and energy dissipation capacity values of glulam wooden beams have increased. However, displacement ductility ratio values have decreased in the experimental specimens strengthened without anchorage CFRP strips. The ultimate load capacity, initial stiffness, and energy dissipation capacity values of glulam wooden beam test specimens, which were strengthened by adhering 1500 mm long CFRP strip, was calculated 18%, 12%, and 25% greater than the test specimens strengthened with 1200 mm long CFRP strips, respectively in average. The ultimate load capacity, initial stiffness, and energy dissipation capacity values of glulam wooden beam test specimens strengthened by adhering 1200 mm long CFRP strip to the tension surface was calculated 9%, 20%, and 26% greater than the test specimens strengthened with 900 mm long CFRP strips, respectively in average. Examining the test results revealed that an increase in the bonding length of the CFRP strips adhered to the lower tensile surface of glulam wooden beams for strengthening has positively affected the general load-displacement behavior of the test specimens under the effect of bending loading. It also increased all the values of the ultimate load capacity, initial stiffness, and energy dissipation capacity. However, the extension of the bonding length of the CFRP strip negatively affected the displacement ductility ratios of the test specimens strengthened without anchorage. The extension of the bonding length of the adhered CFRP strip did not have a significant effect on the displacement ductility ratios of the test specimens strengthened with anchorage CFRP strips. In general, the displacement ductility ratios of the test specimens strengthened with anchorage CFRP strips are lower than the test specimens reinforced without anchorage CFRP strips. The increase in the bonding length of the CFRP strips adhered to the lower tensile surface of glulam wood beams for strengthening has increased the adhesion surface area, and the adhesion surface has more shear stress. With the increase in the adhesion surface area, CFRP strips have been able to carry more axial tensile force by carrying more shear stress. CFRP strips have been more effective in carrying the axial tensile force generated by the bending effect of CFRP strips in glulam wooden beams. However, the increase in the bonding length of the CFRP strips and the increase to the level can carry more axial tensile force. This application decreased ductility in the test specimens and caused a more brittle behavior, the more shear stress carried caused the stress concentration on the CFRP strip to increase.

The differences applied in the strengthening method for the experimental specimens have also been effective on the failure mechanisms and damage distribution resulting. In the strengthening details, using CFRP fan-type anchorages in the endpoints of CFRP strips adhered to bottom tension surfaces of glulam wooden beams has created the greatest effect on the failure mechanisms of the experimental specimens. Specimens 2, 3, and 4, which is strengthened without anchorage CFRP strips, were debonded from a region close to the midpoint of the beam. Then CFRP strip separated from the beam surface and failure with the bending cracks in the glulam wooden beam and fracture in the wooden beam body at the bottom surface lamina. In the test specimens with CFRP fan type anchorage in the CFRP strip endpoints, the CFRP strip was not separated from the wooden beam bottom tension surface, and with the effect of increased loading and bending moment, the test specimens failure with the effect of the bending crack in the wood beam body at the bottom surface lamina after breaking through a region close to the midpoint of the CFRP strip beam.

Numerical analysis

In this section, numerical analysis of the finite element models is utilized by ABAQUS software (ABAQUS Users’ Manual, 2015) to verify the results obtained in the experimental part of the study. Because many material models and analysis characteristics can be defined in ABAQUS finite elements software, it is commonly used by engineers and researchers in recent years. So, the behavior of the various materials and structural members can be obtained after numerical analysis. First of all, finite element models of the timber beams are generated by three-dimensional, eight-node linear brick, hexahedron element type (C3D8R) in the software. Section sizes and length of the beams with anchorage configurations are taken in the same way of experimental study. Besides, strict attention is paid while modeling CFRP strips. Thus, the cohesive zone model of the software is utilized to model the interface between the beam and CFRP. In the final step of the modeling phase, support conditions are defined with boundary conditions at both ends of the specimens just as the experimental part of the study.

After generating finite element models of the specimens, material characteristics are defined in the software. Modeling the manner of the correct material is the basis of getting accurate results from the numerical analysis. As timber is an organic material, its modeling is very complex. The definition of strength and stiffness for different material directions is required due to the anisotropy of timber. Furthermore, behaviors in tension and compression have to be modeled differently. Timber is inhomogeneous with various growth defects. That means that physical and mechanical properties in many cases have to be defined at an element level or cross-section level, rather than at the global material level (Thelandersson, 2003). The mentioned properties of timber were taken into account during material modeling. Not to make the model to a complex, timber was considered to be orthotropic material. Nine independent constants (three moduli of elasticity, three shear modulus, and three Poisson’s ratios) were used to describe the mechanical behavior of timber. A linear-elastic relationship defined the stress-strain behavior of the timber in tension, while a linear elastic-perfectly plastic relationship was used timber in compression. The constitutive low for timber (Figure 6) can be expressed by equations (1)–(3).

σ_{w, t} = E_{w} \cdot ϵ_{w, t}

(1)

σ_{w, c} = E_{w} \cdot ϵ_{w, c} i f ε_{w, c} \leq ε_{w, c y}

(2)

σ_{w, c} = σ_{w, cy} if ε_{w, c} > ε_{w, cy}

(3)

where: σ_w,t, and σ_w,c are the timber tensile and compressive stress; E_w is timber elasticity modulus; ε_w,t, and ε_w,c are tensile and compressive strain in timber; and ε_w,cy is strain value at yield stress σ_w,cy. The expected plastic behavior of top lamination in the compressive zone was modeled using the theory of anisotropic plasticity. Hill’s criterion for orthotropic materials was used as a condition for transition to the plastic state. It represents a generalized version of von Mises’ yield criterion, which considers the anisotropy of the strength of material (Abrate, 2008). Normal compressive yield stresses for the three orthogonal directions, and yield shear stresses in the three shear planes were declared to satisfy the criterion.

Figure 6.

Constitutive law for timber and CFRP.

Linear elastic material models are used to assign material properties such as density, elastic and shear modulus, strength values in compression, and tension in the CFRP constitutive material model (Figure 6). Besides, the surface to surface contact behavior is defined for the geometries of the related sections. The contact between the specimen and CFRP is supposed to be perfect contact without any separation. Before performing numerical analysis of test specimens, finite element models of the test specimens are separated into small pieces to obtain more accurate results. After performing mesh operation for Specimen 1, the finite element size is decided as 10 mm for all geometries. Time interval (Δt) is another important parameter in the numerical analysis. This value is defined as 0.01 s in the software. Afterward, numerical analysis is performed under the effect of three-point bending by considering the changes in CFRP strips and anchorage configurations. The analysis is continued until reaching the failure damage situation for each test specimen. Analysis results are considered to be useful in the verification of experimental results. Load-midpoint displacement curves are also obtained after numerical analysis in the software. Two of these curves are shown in Figure 7 for Specimen-7 and Specimen-8 as examples. The numerical energy dissipation capacities of the specimens are calculated by considering the area under these curves. As a result, ultimate load capacities, displacement at ultimate load, displacement at failure load, and energy dissipation capacity values of the specimens are comparatively given in Table 5. Average ratios of experimental to numerical values are also calculated to reveal the relationship between both results. When the experimental and numerical analysis results are investigated, it’s seen that a good relationship is established between both studies. The average ratio of ultimate load capacity values is calculated as 1.06 between experimental and numerical results. So, a 6% difference is obtained due to the results of all test specimens. Average ratio increases to 1.08 when displacement values at ultimate load are compared. The ratios of displacement values at failure load vary between 1.00 and 1.14, and the average ratio is determined as 1.09. When the error rates between experimental and numerical results are investigated, it’s considered that slight support movements during the experimental study and analysis parameters and material characteristics in the software can be the main reasons.

Figure 7.

Comparison of experimental and numerical load-midpoint displacement curves for two specimens as examples.

Table 5.

Comparison of experimental and numerical results.

Specimen no.	Ultimate load capacity (kN)			Displacement at ultimate load (mm)			Displacement at failure load(mm)			Energy dissipation capacity(kN mm)
Specimen no.	Exp.	Num.	Ratio^a	Exp.	Num.	Ratio^b	Exp.	Num.	Ratio^c	Exp.	Num.	Ratio^d
1	20.53	18.86	1.09	48.46	44.47	1.09	48.46	46.51	1.00	620.68	682.67	0.90
2	31.58	28.79	1.10	46.58	42.16	1.10	46.58	42.16	1.10	869.12	829.49	1.04
3	25.60	24.56	1.04	37.58	38.54	0.98	46.92	42.13	1.11	832.32	804.14	1.03
4	22.81	22.13	1.03	31.67	33.47	0.95	42.28	38.57	1.10	643.65	617.59	1.04
5	38.83	35.42	1.10	43.08	40.92	1.05	52.71	46.34	1.14	1339.34	1174.64	1.14
6	36.72	33.61	1.09	42.45	37.74	1.12	42.45	41.69	1.02	849.97	1028.00	0.82
7	33.55	31.92	1.05	37.46	33.18	1.13	37.46	33.18	1.13	835.15	726.29	1.15
8	52.19	48.77	1.07	41.92	38.56	1.09	41.92	38.56	1.09	1447.69	1284.30	1.13
9	42.04	40.82	1.03	39.90	35.81	1.11	44.55	39.06	1.14	1267.31	1184.68	1.07
10	40.49	38.53	1.05	38.45	31.37	1.23	38.45	34.57	1.11	861.75	979.57	0.88
Average			1.06			1.08			1.09			1.02

Ratio of experimental ultimate load capacity values to numerical results.

Ratio of experimental displacement at ultimate load values to numerical results.

Ratio of experimental displacement at failure load values to numerical results.

Ratio of experimental energy dissipation capacity values to numerical results.

In the last step of the numerical analysis, fracture distributions are obtained from the software under three-point bending. For this purpose, the Hashin damage model of the software is utilized (Khorsandnia et al., 2013). The Hashin damage criteria capture the nonlinear behavior of timber. The Hashin damage model takes account of the interaction between stress components to evaluate different failure modes. The Hashin model was developed initially as failure criteria for unidirectional polymeric composites, and hence, application to other laminate types or non-polymeric composites represents an approximation. This model successfully presents the damage behavior of the finite element models and yields satisfactory results under different loading types. The 2D Hashin failure criteria employed in this study are used to predict the onset of damage. The damage evolution law is based on the energy dissipated during the damage process and linear material softening. Fracture distributions are investigated when the load is completely applied to the specimens. It is seen that fracture distributions exhibit similar behavior with damage development of the test specimens in the experimental study. Besides, the distributions are generally accumulated around static loading point and expand to the supports as given in Figure 5.

Conclusion

In this study, a reinforcement method has been developed with CFRP strips to increase the performance of wooden glulam beams under the effect of flexure loading and to improve the load-displacement behavior. In addition to increasing the maximum load-carrying capacity and initial stiffness values of glulam wooden beams in the developed strengthening method, it is also aimed not to affect displacement-ductility ratios negatively and to increase energy dissipation capacity values. In order to increase the load-carrying capacity, initial stiffness, and energy dissipation capacity as targeted in the scope of this study, the application of CFRP fan type anchorage at the endpoints of the CFRP strips is an innovative reinforcement detail which is applied in this research. A verified numerical analysis model was created by using the results obtained from the experimental study. The results obtained by using ABAQUS finite element software are compared with the experimental results, and it is interpreted to what extent successful computer simulation results can be achieved. The results obtained from the study are presented below as items.

The strengthening method applied with CFRP strips positively affected the maximum load-carrying capacity, initial stiffness, displacement ductility ratios, and energy consumption capacity values of glulam wooden beams by increasing them significantly.

As an innovative method, proposed CFRP fan-type anchors in the endpoints of the CFRP strips in the strengthening details positively affected and increased the maximum load-carrying capacity, initial stiffness, and energy consumption capacity values of the test specimens. Compared to glulam wooden beams strengthened only with CFRP strip, the test specimens strengthened with anchorage CFRP strips increased the maximum load-carrying capacity, initial stiffness, and energy dissipation capacity without strengthening reference test specimen, and substantially increased the performance of the strengthened glulam beams.

The CFRP fan type anchorage application used in CFRP strip endpoints negatively affected displacement-ductility ratios. Also, the displacement-ductility ratio increased in test specimens strengthened with an anchorage at endpoints of CFRP strips compared to experimental specimens strengthened only with CFRP strips.

In the developed strengthening method, as the length of CFRP strips adhered to the tensile surface of glulam wooden beams in the developed reinforcement method increase, the maximum load-carrying capacity, initial stiffness, and energy dissipation capacity values of the test specimens increased. Although, displacement-ductility ratio values decreased in the experimental specimens reinforced only with CFRP strips.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Özgür Anıl

References

ABAQUS User’s Manual, Version 6.12, SIMULIA, Dassault Systèmes Simulia Corp., 2015.

Abrate

(2008) Criteria for yielding or failure of cellular materials. Journal of Sandwich Structures and Materials 10(1): 5–51.

Alhayek

Svecova

(2012) Flexural stiffness and strength of GFRP-reinforced timber beams. Journal of Composites for Construction ASCE 16(3): 245–252.

Ambrisi

Focacci

Luciano

(2014) Experimental investigation on flexural behavior of timber beams repaired with CFRP plates. Composites Structures 108: 720–728.

Andre

(2006) Fibres for strengthening of timber structures. Research report, Lulea University of Technology, Sweden.

Andor

Lengyel

Polgár

, et al. (2015) Experimental and statistical analysis of spruce timber beams reinforced with CFRP fabric. Construction and Building Materials 99: 200–207.

Borri

Corradi

Grazini

(2005) A method for flexural reinforcement of old wood beams with CFRP materials. Composites Part B: Engineering 36: 143–153.

Borri

Corradi

Speranzini

(2013) Reinforcement of wood with natural fibers. Composites Part B: Engineering 53: 1–8.

Cabrero

Heiduschke

Haller

(2010) Analytical assessment of the load carrying capacity of axially loaded wooden reinforced tubes. Composite Structures 92(12): 2955–2965.

10.

Conde

MJM

Liñán

Hita

(2015) Bending and shear reinforcements for timber beams using GFRP plates. Construction and Building Materials 96: 461–472.

11.

Corradi

Righetti

Borri

(2015) Bond strength of composite CFRP reinforcing bars in timber. Materials 8: 4034–4049.

12.

Fiorelli

Dias

(2011) Glulam beams reinforced with FRP externally-bonded: Theoretical and experimental evaluation. Materials and Structures 44: 1431–1440.

13.

Fossetti

Minafò

Papia

(2015) Flexural behaviour of glulam timber beams reinforced with FRP cords. Construction and Building Materials 95: 54–64.

14.

Franke

Harte

(2015) Failure modes and reinforcement techniques for timber beams – State of the art. Construction and Building Materials 97: 2–13.

15.

Gentile

Dagmar

Rizkalla

(2002) Timber beams strengthened with GFRP bars: Development and applications. Journal of Composites for Construction 6(1): 11–20.

16.

Gezer

Aydemir

(2010) The effect of the wrapped carbon fiber reinforced polymer material on fir and pine woods. Materials & Design 31: 3564–3567.

17.

Gilfillan

Gilbert

Russell

(2001) Enhancement of the structural performance of home-grown sitka spruce using carbon fibre reinforced polymer. The Institution of Structural Engineers 79(8): 23–28.

18.

Gilfillan

Gilbert

Patrick

GRH

(2003) The use of FRP composites in enhancing the structural behaviour of timber beams. Journal of Reinforced Plastics and Composites 22(15): 1373–1388.

19.

Haller

Heiduschke

Putzger

, et al. (2015) Compressed laminated wood. Bautechnik 92(1): 28–35.

20.

Hernandez

Davalos

Sonti

, et al. (1997) Strength and stiffness of reinforced yellow-poplar glued laminated beams. US Department of Agriculture, Forests Products Laboratory Res. Pap FPL-RP-554.

21.

Issa

Kmeid

(2005) Advanced wood engineering. Construction and Building Materials 19: 99–106.

22.

Jasienko

Nowak

(2014) Solid timber beams strengthened with steel plates experimental studies. Construction and Building Materials 63: 81–88.

23.

Johnsson

Blanksvard

Carolin

(2006) Glulam members strengthened by carbon fiber reinforcement. Materials and Structures 40: 47–56.

24.

Kasal

Pospisil

Jirovsky

, et al. (2004) Seismic performance of laminated timber frames with fiber-reinforced joints. Earthquake Engineering Structural Dynamics 33(5): 633–646.

25.

Khelifa

Auchet

Meausoone

, et al. (2015). Finite element analysis of flexural strengthening of timber beams with carbon fibre-reinforced polymers. Engineering Structures 101: 364–375.

26.

Khelifa

Celzard

(2015) Flexural strengthening of finger-jointed Spruce timber beams with CFRP. Journal of Adhesion Science and Technology 29(19): 2104–2116.

27.

Khorsandnia

Valipour

Crews

(2013) Nonlinear finite element analysis of timber beams and joints using the layered approach and hypoelastic constitutive law. Engineering Structures 46: 606–614.

28.

Kim

Harries

(2010). Modeling of timber beams strengthened with various CFRP composites. Engineering Structures 32: 3225–3234.

29.

Tsai

Wei

, et al. (2014) A study on wood beams strengthened by FRP composite materials. Construction and Building Materials 62: 118–125.

30.

Xie

Tsai

(2009) Enhancement of the flexural performance of retrofitted wood beams. Construction Building and Materials 23: 411–422.

31.

Luca

Marano

(2011) Comparison of unreinforced and reinforced glulam timber with steel bars. European Journal of Technology and Advances in Engineering and Research 2: 45–54.

32.

Ling

Geng

, et al. (2015) Study on flexural behaviour of glulam beams reinforced by Near Surface Mounted (NSM) CFRP laminates. Construction and Building Materials 91: 23–31.

33.

Mark

(1961) Wood–aluminium beams within and beyond the elastic range. Part 1: Rectangular sections. Forest Products Journal 11(10): 477–484.

34.

McConnell

McPolin

Taylor

(2014) Post-tensioning of glulam timber with steel tendons Construction and Building Materials 73: 426–433.

35.

Naghipour

Nematzadeh

Yahyazadeh

(2011) Analytical and experimental study on flexural performance of WPC-FRP beams. Construction and Building Materials 25: 829–837.

36.

Negrao

JHJO

Balseiro

Faria

(2011) Use of GFRP laminates for strengthening or rehabilitation of timber beams. In: SHATIS’11-international conference on structural health assessment of timber structures, Lisbon, Portugal.

37.

Ogawa

(2000) Architectural application of carbon fibers – Development of new carbon fiber reinforced glulam. Carbon 38: 211–226.

38.

Raftery

Harte

(2011). Low grade glued laminated timber reinforced with FRP plate. Composites Part B: Engineering 42: 724–735.

39.

Raftery

Harte

(2013) Nonlinear numerical modeling of FRP reinforced glued laminated timber. Composites Part B: Engineering 52: 40–50.

40.

Raftery

Kelly

(2015) Basalt FRP rods for reinforcement and repair of timber. Composites Part B: Engineering 70: 9–19.

41.

Riberio

Jesus

AMP

Lima

, et al. (2009) Study of strengthening solutions for glued-laminated wood beams of maritime pine wood. Construction and Building Materials 23: 2738–2745.

42.

Schober

Harte

Kliger

, et al. (2015) FRP reinforcement of timber structures. Construction and Building Materials 97: 106–118.

43.

Sena-Cruz

Jorge

Branco

, et al. (2013) Bond between glulam and NSM CFRP laminates. Construction and Building Materials 40: 260–269.

44.

Thelandersson

(2003) Introduction: Safety and serviceability in timber engineering. In: Thelandersson

Larsen

(eds.). Timber Engineering. Chichester, England: John Wiley and Sons, Ltd., pp.171–176.

45.

Triantafillou

Deskovic

(1992) Prestessed FRP sheets as external reinforcement of wood members. Journal of Structural Engineering ASCE 118(5): 1270–1284.

46.

Valipour

Crews

(2011) Efficient finite element modeling of timber beams strengthened with bonded fiber reinforced polymers. Construction and Building Materials 25: 3291–3300.

47.

Yang

Liu

Xiong

(2013). Flexural behavior of wood beams strengthened with HFRP. Construction and Building Materials 43: 118–124.

48.

Yeboah

Taylor

McPolin

, et al. (2013) Pull-out behaviour of axially loaded basalt fibre reinforced polymer (BFRP) rods bonded perpendicular to the grain of glulam elements. Construction and Building Materials 38: 962–969.