Study on mechanical properties of poplar timber bending members of ancient buildings strengthened with CFRP sheets embedded

Abstract

The reduction of the section area of timber beams in ancient buildings leads to the decline of their flexural mechanical properties. In order to study the effect of the reinforcement method with near-surface mounted CFRP sheets, three groups of 14 scaled poplar timber beam specimens were designed for static loading tests, and the strengthening effects of the new members replacing the damaged members and the concealed CFRP sheets strengthening the damaged members were compared. The results show that the ultimate flexural capacity and flexural stiffness of the strengthened timber beams are significantly improved compared with the simulated damaged beams; the bending performance of the optimum strengthened timber beam is equivalent to that of the undamaged beam. The distribution of the section strain of the strengthened timber beam along the height of the beam section basically conforms to the plane section assumption. A formula for calculating the ultimate flexural capacity of poplar timber beams strengthened with CFRP sheets is proposed. The calculated results are in good agreement with the experimental results.

Keywords

Concealed reinforcement with CFRP sheets poplar timber beams flexural load capacity flexural stiffness

1. Introduction

Chinese ancient architecture timber structure components are mostly taken from local timber species, in northern China, for example, often use pine, poplar, elm, cypress, etc. [1, 2] Among them, small leaf poplar because of the wood is light and soft, straight texture, easy processing, mostly seen in ancient architecture beam, column elements. However, as a biological material, poplar bending members are prone to deformation, cracks, decay and insect infestation over a long period of time due to natural and man-made factors, which affects the safety of the building structure and causes the loss of valuable information. Carbon Fiber-reinforced Polymer (CFRP) has attracted wide attention in the field of architectural heritage conservation due to its excellent properties such as light weight, high strength, fatigue resistance, corrosion resistance and strong designability. The method of externally bonded reinforcement (EBR) on the surface of timber members only considers the improvement of mechanical properties, ignoring the damage to the color paintings and inscriptions on the surface of timber members. This not only leads to the loss of the value information of ancient buildings, but also violates the basic principles of cultural heritage conservation. Therefore, for the timber beam members of ancient buildings, the near-surface mounted (NSM) method of grooving on the surface, embedding CFRP sheets, and filling the gaps in the groove with structural adhesive has more advantages and practical significance. It can protect the materials and historical information of components to the greatest extent, and better meet the needs of architectural heritage conservation than EBR method.

In recent years, scholars at home and abroad have applied NSM FRP technique to carry out a lot of research on the reinforcement of timber members of ancient buildings. Yeboah et al. [3] passed the bending test on 20 timber beamsï¼ŒNSM BFRP and GFRP reinforced timber beams were studied. The flexural behaviour was improved by FRP reinforcement. The proposed theoretical model for the ultimate strength of NSM FRP strengthened timber beams is assessed on the basis of the test results and collated data, showing a good comparison between the experimental and theoretical results. Bakalarz et al. [4] used NSM CFRP sheets method to strengthen beams made of full-size laminated veneer lumber (LVL). The reinforcement mainly affected the enhancement of the maximum bending moment values evaluated at the points of application as having concentrated forces of 32% and 24% in comparison to the unreinforced elements. The reinforced elements were characterized by a greater variation in failure mode, resulting from tension, compression or lateral torsional buckling. Chun et al. [5] In order to investigate the flexural performance of CFRP sheets and tendons reinforced fir and pine beams by the NSM method, flexural tests were conducted on 20 specimens, the results showed that the flexural load capacity and stiffness of the specimens reinforced with CFRP sheets and bars were improved compared with those without reinforcement. Xu et al. [6] conducted an experimental study on the flexural performance of timber beams reinforced with embedded CFRP sheets, and the results showed that the damage of timber beams reinforced with embedded CFRP sheets (plates) originated from wood defects at the tensile and compressive edges, and their ductility was significantly improved; the flexural load capacity of the specimens was significantly increased, with an average increase of 39.3%. Chun et al. [7] conducted an experimental study on the flexural properties of timber beams reinforced with NSM CFRP sheets, and the results showed that the flexural load capacity of timber beams was significantly improved, with the increase of 9.1% $\sim$ 35.1% and 19.8% $\sim$ 37.4% for pine and fir specimens, respectively. Finally, the calculation equations for the flexural load capacity of rectangular beams of pine and fir reinforced with NSM CFRP sheets were established. Zhu et al. [8] studied the flexural performance of the damaged timber beams of the concealed reinforced ancient buildings with FRP sheets. The results showed that the ultimate flexural capacity and flexural rigidity of the reinforced beams were significantly improved compared with the unreinforced beams, and the flexural performance of the optimal reinforced timber beams was equivalent to that of the undamaged beams.

In summary, the research on the reinforcement of poplar bending members for ancient buildings in North China is still incomplete. In this paper, experimental research and theoretical analysis are conducted on the flexural performance of poplar beams reinforced with NSM CFRP sheets, and corresponding conclusions and suggestions are made based on the results.

2. Test materials and methods

2.1 Materials

The simon poplar commonly used in timber components such as beams, purlins and braces of ancient buildings in Beijing, China, was selected as the test material. The timbers were selected from the same batch of new timber and tested by the German IML microdrill resistance meter and the Hungarian Fakopp stress wave meter. The mechanical properties of the timber members were deduced from the wave resistance modulus $\rho$ v ${}^{2}$ [9], whereby the parts of similar timber properties were selected for the test pieces, while avoiding knots and defects on the The test specimens are then selected for their similarity in wood properties. Referring to the Chinese national standard for physical and mechanical properties of wood (GB/T1938-2009 and GB/T1936.1-2009) , the static flexural strength of poplar was measured at 76.8 MPa, the modulus of elasticity at 10,012 MPa, the compressive strength along the grain at 38.7 MPa, and the moisture content at 10.6%. The CFRP sheets were supplied by Carbone Composites (Tianjin) Co. 160 GPa elongation of 1.7%. The carbon sheet adhesive is the domestic Carbon-CEPR-A/B carbon sheet adhesive with a standard tensile strength of 43 MPa and a tensile modulus of elasticity of 2410 MPa.

2.2 Experimental design

In order to investigate the effect of the amount and location of reinforcement of CFRP sheets on the flexural mechanical properties of reinforced beams, the poplar specimen types were divided into a total of three categories: undamaged beams, simulated salvage beams and concealed reinforced beams with CFRP sheets, as shown in Table 1.

Table 1
Specimens design

Type of test piece	Specimen number	Slot size/mm	Reinforcement position and quantity	Reinforcement ratio (%)
Undamaged	PSB01	–	–	–
beams	PSB02	–	–	–
Simulated	PSB03	10 $\times$ 6 $\times$ 900	–	–
damaged	PSB04	10 $\times$ 6 $\times$ 900	–	–
beams	PSB05	6 $\times$ 20 $\times$ 900	–	–
	PSB06	6 $\times$ 20 $\times$ 900	–	–
	PSB07	6 $\times$ 20 $\times$ 900	–	–
	PSB08	6 $\times$ 20 $\times$ 900	–	–
Concealed reinforced	PSB09	10 $\times$ 6 $\times$ 900	Side, one sheet is embedded	0.74
beams with	PSB10	10 $\times$ 6 $\times$ 900	Side, one sheet is embedded	0.74
CFRP sheets	PSB11	6 $\times$ 20 $\times$ 900	Bottomã€one sheet is embedded	0.74
	PSB12	6 $\times$ 20 $\times$ 900	Bottom, one sheet is embedded	0.74
	PSB13	6 $\times$ 20 $\times$ 900	Bottom, two sheets are embedded	1.48
	PSB14	6 $\times$ 20 $\times$ 900	Bottom, two sheets are embedded	1.48

The timber beam specimens were made according to the data on medical devices for four rafters of the timber structure in Song Dynasty [10] and were modelled at a scale of 1:2.5, with dimensions of 60 mm (W) $\times$ 90 mm (H) $\times$ 1100 mm (L). For undamaged beams PSB01 and 02, the specimens are intact. For simulated stumped beams PSB03 and 04, a slot is cut on the left and right side of the specimen with dimensions of 10 mm (W) $\times$ 6 mm (H) $\times$ 900 mm (L); for simulated stumped beams PSB05 and 06, a slot of 6 mm (W) $\times$ 20 mm (H) $\times$ 900 mm (L) is cut along the centre line at the bottom of the specimen; for simulated stumped beams PSB07 and 08, two slots are symmetrically cut on the bottom of the timber beam along the center line, left and right, and the dimensions of both slots are 6 mm (width) $\times$ 20 mm (height) $\times$ 900 mm (length). The above specimens are used to simulate the weakening of the section, such as cracking, surface defects and other types of damage; for CFRP reinforced beams PSB09 to PSB14, a 2.0 mm thick CFRP sheet is embedded in the slotted position of the corresponding simulated damaged beam and the slot is filled with carbon adhesive. The voids in the slots were filled with carbon adhesive.

Figure 1.

Testing equipments.

The production process of the test specimen is as follows: 1. Use woodworking tools to shape the length, width and height of the timber beam according to the design drawings, and plane the surface; 2. Use power tools to open different size slots on the bottom or side of the reinforced beam according to the test program; 3. Clean the wooden slots with air guns and brushes; 4. Embed the cut CFRP sheets and fill the slots with carbon glue, and fix the surface with straps; 5. Place the reinforced test specimen on the horizontal ground at 25 ${{}^{\circ}}$ C for 7 days; 6. Paste the strain gauges on the surface of the test specimen at the corresponding position according to the test program.

2.3 Test methodology

The test was carried out at the Engineering Mechanics Experimental Centre of Beijing University of Technology, according to the Standard for Test Methods for Timber Structures (GB/T50329-2012), using a QBD100 universal mechanical testing machine for four-point bending test, the load was transferred through a distribution beam, the loading rate was approximately 5 mm/min, and the loading time for each specimen was 8 to 12 minutes, as shown in Fig. 1. The test was carried out using a German IMC 8-channel dynamic strain gauge with a sampling frequency of 20 Hz. The range of the load cell was 50 kN, accuracy class I, with a minimum resolution of $\pm$ 1 N. The test was carried out using a displacement gauge with a range of 30 mm, accuracy class I, with a minimum resolution of 10 ${}^{-3}$ mm.

3. Test results and analysis

3.1 Damage phenomena

Figure 2.

Failure modes of specimens.

Poplar undamaged specimens PSB01 and 02, suffered damage from the pulling off of the wood fibres from 1/2 of the beam section to the bottom surface. Poplar simulated stumpage specimens PSB03-PSB08 all suffered tensile damage due to severe splitting at the span edge at the bottom of the beam and most of the span deflection was not recoverable after unloading. For the poplar reinforced beams: specimens PSB09 and PSB10 with slabs embedded in the side of the beam and specimens PSB11 and PSB12 with one slab embedded in the bottom of the beam, all suffered damage from the pulling off of the wood fibres from 1/2 of the beam section to the bottom surface, with jagged cracks and a broken and partially detached glue surface. Specimens PSB13 and PSB14 with two panels embedded at the bottom of the beam had the timber pulled off at the span edge at the bottom of the beam, eventually causing damage.

3.2 Load-displacement curves and flexural mechanical properties

The load-displacement curves are shown in Fig. 3. Although the same batch of new timber with similar properties was selected for the test, it is still difficult to avoid the influence of the dispersion of the mechanical properties of wood on the test results. In the case of large differences in load capacity, the values should also be taken in conjunction with the actual condition of the specimen.

Table 2
Main experimental results

Type of test piece	Specimen number	Ultimate bearing capacity/kN (P)	Average value	EI/kN $\cdot$ m ${}^{2}$	Average value
Undamaged beams	PSB01	24.83	24.59	20.71	22.49
	PSB02	24.35		24.26
Simulated damaged beams	PSB03	13.87	13.73	11.94	12.67
	PSB04	13.58		13.39
	PSB05	13.75	14.69	14.58	16.27
	PSB06	15.62		17.96
	PSB07	11.89	11.50	14.70	15.70
	PSB08	11.10		16.70
Concealed reinforced	PSB09	21.10	20.81	20.93	20.39
beams with CFRP sheets	PSB10	20.51		19.85
	PSB11	25.41	25.26	15.15	17.15
	PSB12	25.10		19.15
	PSB13	33.52	31.08	17.47	20.88
	PSB14	28.64		24.28

As seen in Table 2 and Fig. 3, (1) Compared with the unreinforced simulated damaged beams, the ultimate load capacity of the laterally reinforced specimens increased by 58.8% and the flexural stiffness by 60.9% on average; the ultimate load capacity of the specimens reinforced with one CFRP sheets embedded in the bottom side increased by 71.9% and the flexural stiffness by 5.4% on average; the ultimate load capacity of the specimens reinforced with two CFRP sheets embedded in the bottom side increased by 170.3% and the flexural stiffness by 33.0% on average. The average increase in ultimate load carrying capacity of the specimens with two CFRP sheets at the bottom was 170.3%, and the increase in flexural stiffness was 33.0%. (2) In the poplar reinforced beams, the ultimate load capacity of the bottom reinforced specimens increased significantly and exceeded that of the undamaged beams, except for the side reinforced specimens which were slightly lower than the undamaged specimens. (3) The ductility and flexural stiffness of the poplar-reinforced beams were significantly higher than those of the simulated damaged beams.

Figure 3.

Loading-strain curves of specimens.

3.3 Stress-strain curves

Five strain gauges are evenly arranged along the height direction of the timber beam section with a spacing of 22.5 mm, see Fig. 1(a). Taking undamaged beam PSB02, simulated damaged beam PSB04 and reinforced beam PSB10 as examples, the strain distribution along the height direction of the span mid-section was analysed: when the strain was negative, the section was in compression, i.e. the section was located above the neutral layer; when the strain was positive, the section was in tension, i.e. the section was located below the neutral layer. From Fig. 4(a) and (b), it can be seen that the strains along the height of the section in the span of the beam specimen are linearly distributed, which basically conforms to the assumption of flat section.

Figure 4.

Strain profile at mid-span cross-section of typical strengthened beams.

4. Theoretical calculation model for reinforced timber beams

In the theoretical derivation of the model for calculating the flexural load capacity of reinforced timber beams, the following basic assumptions are made in accordance with references [7, 11].

(1)
The strain distribution in the cross-section of a timber beam after bending remains in accordance with the flat section assumption.
(2)
The timber is of uniform material, free from defects such as knots, wormholes and cracks.
(3)
The modulus of elasticity of wood in tension, compression and bending is the same.
(4)
Wood exhibits linear elasticity when subjected to tension and compression.
(5)
Carbon fibre sheets using a linear elastic stress-strain relationship.
(6)
The bond between the embedded fibreboard and the timber and the carbon board adhesive is solid and reliable until the bending load limit state is reached, no slippage occurs and the deformation is always coordinated.

The test moment can be obtained from Eq. (1) , where is the measured ultimate load capacity of the beam and is the horizontal distance from the loading point to the support point.

$\displaystyle M_{e}=\frac{P}{2}$ (1)

The cross-section consisting of three materials – wood, carbon sheet adhesive and carbon fibre board – is converted into an equivalent cross-section of a single material by the transformed cross-section method and then solved using the method of analysing homogeneous material beams.

Figure 5.
Cross-section of strengthened beams analysis chart.

From Eqs (2) to (6), the bending moments for each of the three materials – wood, carbon glue and fibreboard – can be found as $M_{w}$ , $M_{g}$ , and $M_{p}$ respectively. The calculated bending moments for $M_{c}$ are shown in Eq. (7). Where: $h_{1}$ is the height of the section, $b_{1}$ is the width of the section, $b_{2}$ is the width of the groove, $h_{2}$ is the depth of the wood groove, $b_{3}$ is the width of the fibreboard and $h_{3}$ is the height of the fibreboard, as shown in Fig. 5. $y\ y_{c}$ is the height of the section in the compression zone, $n_{1}$ is the modulus of elasticity ratio between the gel and the wood, $n_{2}$ is the modulus of elasticity ratio between the fibreboard and the wood and $E$ is the modulus of elasticity of the wood. $\varepsilon$ is the strain of the beam, obtained by experimentally measuring the strain at different section heights and fitting a straight line. (Note: the upper surface of the timber is used as the $x z$ plane)

$\displaystyle y_{c}=\frac{\frac{1}{2}b_{1}h^{2}_{1}+(n_{1}-1)b_{2}h_{2}\left(h% _{1}-\frac{h_{2}}{2}\right)+(n_{2}-n_{1})b_{3}h_{3}\left(h_{1}-\frac{h_{3}}{2}% \right)}{b_{1}h_{1}+(n_{1}-1)b_{2}h_{2}+(n_{2}-n_{1})b_{3}h_{3}}$ (2) $\displaystyle\varepsilon=ky+c$ (3) $\displaystyle M_{w}=\int^{h_{1}-h_{2}}_{0}10^{-6}\varepsilon\cdot E\cdot b_{1}% \cdot(y-y_{c})dy+\int^{h_{1}}_{h_{1}-h_{2}}10^{-6}\varepsilon\cdot E\cdot(b_{1% }-b_{2})\cdot(y-y_{c})dy$ (4) $\displaystyle M_{g}=\int^{h_{1}-h_{3}}_{h_{1}-h_{2}}10^{-6}\varepsilon\cdot E% \cdot n_{1}b_{2}\cdot(y-y_{c})dy+\int^{h_{1}}_{h_{1}-h_{3}}10^{-6}\varepsilon% \cdot E\cdot n_{1}(b_{2}-b_{3})\cdot(y-y_{c})dy$ (5) $\displaystyle M_{p}=\int^{h_{1}}_{h_{1}-h_{3}}10^{-6}\varepsilon\cdot E\cdot n% _{2}b_{3}\cdot(y-y_{c})dy$ (6) $\displaystyle M_{c}=M_{w}+M_{g}+M_{p}s$ (7)

Based on the above calculation model, the calculated values of bending moments for each reinforced beam were derived and compared with the test values, as shown in Table 3, which shows that the ratio of calculated values to test values ranged from 0.84 to 1.03, which is close to the calculated values and test values, verifying the accuracy of the calculation model. In addition, according to the reference [8], the bending moment of timber and CFRP sheets in the reinforced beam accounts for 86.4% and 12.7% of the total bending moment value, and the bending moment of adhesive is only 0.9% of the total bending moment value, so the bending moment of adhesive can be ignored. At the same time, based on the theoretical calculation model obtained from the scale test, it is still necessary to make corrections and improvements through the full-scale beam reinforcement test data.

Table 3
Comparison between predicted load capacity and test results

Specimen number $M_{g}$ (kN $\cdot$ m) $M_{c}$ (kN $\cdot$ m) $M_{g}/M_{c}$

PSB09 3.46 3.89 0.89

PSB10 4.02 4.71 0.85

PSB11 3.38 3.28 1.03

PSB12 3.45 4.11 0.84

PSB13 1.32 1.35 0.98

PSB14 4.38 4.59 0.95

5. Conclusion

Specimen number	$M_{g}$ (kN $\cdot$ m)	$M_{c}$ (kN $\cdot$ m)	$M_{g}/M_{c}$
PSB09	3.46	3.89	0.89
PSB10	4.02	4.71	0.85
PSB11	3.38	3.28	1.03
PSB12	3.45	4.11	0.84
PSB13	1.32	1.35	0.98
PSB14	4.38	4.59	0.95

(1)

The distribution of section strains along the height direction of the beam section for the poplar undamaged beams, the simulated stumpage beams and the reinforced beams are generally in accordance with the flat section assumption.

(2)

The concealed reinforcement by CFRP sheets can effectively improve the flexural mechanical properties of simulated damaged poplar timber beams. The flexural mechanical properties of the reinforced beams can be restored to the level of undamaged beams.

(3)

By calculating the bending moment of the reinforced timber beams, the results are close to the experimental values, so the calculation model has some reference value for the research related to the concealed reinforcement of poplar timber beams of ancient buildings with CFRP sheets.

(4)

The size effect has a great impact on the strength of timber. Since China has not yet established a size adjustment method for the bending strength of the specification timber, the theoretical calculation model based on this test has yet to be revised and improved in combination with the data of the full-scale beam reinforcement test. In addition, in reality, there are many types of damage to the timber beam components of ancient buildings, such as the cracking of the top surface of the beam, or the degradation of the mechanical properties of the timber itself, etc. The effect of concealed reinforcement with CFRP sheets needs further study.

Footnotes

Acknowledgments

Supported by the Scientific Research Initiation Fund of North China University of Technology and Beijing Natural Science Foundation (8232006).

References

Yang

et al., Internal cause analysis of damage of wooden components in danxia temple ancient architectures: Tree species. Wood Research. 2021; 66(2): 297-308.

Yuan

et al., Identification of common wood species of wooden components in ancient buildings based on micro-destructive testing. Linye Kexue/Scientia Silvae Sinicae. 2021; 57(12): 122-131.

Yeboah

. Gkantou, Investigation of flexural behaviour of structural timber beams strengthened with NSM basalt and glass FRP bars. Structures. 2021; (33): 390-405.

Bakalarz

Kossakowski

Tworzewski

. Strengthening of bent LVL beams with near-surface mounted (NSM) FRP reinforcement. Materials. 2020; 13(10): 2350.

Chun

Balen

Pan

. Flexural performance of small fir and pine timber beams strengthened with near-surface mounted carbon-fiber-reinforced polymer (NSM CFRP) plates and rods. International Journal of Architectural Heritage. 2016; 10(1): 106-117.

Zhu

et al., Experimental study on flexural behavior of timber beams strengthened with embedded CFRP bars/sheets. Journal of Building Structure. 2012; (8): 149-156.

Chun

Zhang

. Experimental study on flexural behavior of timber beams strengthened with embedded carbon fiber boards. Journal of Southeast University (Natural Science Edition). 2012; 42(6): 1146-1150.

Zhu

, et al. Study on bending behaviors of ancient building timber beams strengthened with near-surface mounted FRP sheets. Journal of Beijing University of Technology (Natural Science Edition). 2019; 45(2): 160-167.

Wang

et al., Comprehensive Evaluation Method of Historical Timber Structural Building Taking Fujiu Zhou House as an Example. Forests. 2021; 12(9): 1172.

10.

Chen

et al., Degradation laws of hysteretic behaviour for historical timber buildings based on pseudo-static tests. Engineering Structures. 2018; (156): 480-489.

11.

Chun

Zhang

Pan

. Test on flexural performance of wood beams strengthened with embedded carbon fiber reinforcement. Journal of PLA University of Science and Technology (Natural Science Edition). 2013; 14(2): 190-194.

Study on mechanical properties of poplar timber bending members of ancient buildings strengthened with CFRP sheets embedded

Abstract

Keywords

1. Introduction

2. Test materials and methods

2.1 Materials

2.2 Experimental design

Table 1 Specimens design

3. Test results and analysis

3.1 Damage phenomena

Table 2 Main experimental results

Footnotes

Acknowledgments

References

Table 1
Specimens design

Table 2
Main experimental results