Wood-based columns reinforced with fiber-reinforced polymer: A systematic literature review

Abstract

Fiber-reinforced polymer (FRP) is an external reinforcement solution for wooden structures, where several studies have shown its efficiency in maintenance and design. This solution is not yet among the main topics of literature, although its importance justifies new research on this relevant topic for construction. This systematic literature review involves the FRP as reinforcement in wood-based columns, using Engineering Village and Web of Science databases and PRISMA protocol to follow the procedures and ensure the quality of sampling. Reinforcement dispositions and types of assessments were identified so that the literature synthesis can contribute to identifying behavior models. Different methods of reinforcement sizing studied by the literature were synthesized and detailed as to their respective uses. A positive correlation between the reinforcement index and the increasing the load capacity of timber columns were discussed and statistically analyzed.

Full Article

Wood-based Columns Reinforced with Fiber-Reinforced Polymer: A Systematic Literature Review

Pedro I. L. G. Jardim ,^a,b,* Herisson F. dos Santos ,^c Lucas W. A. Mattias ,^d Victor A. De Araujo ,^b,e Júlio C. Molina ,^f and André L. Christoforo ,^b

Fiber-reinforced polymer (FRP) is an external reinforcement solution for wooden structures, where several studies have shown its efficiency in maintenance and design. This solution is not yet among the main topics of literature, although its importance justifies new research on this relevant topic for construction. This systematic literature review involves the FRP as reinforcement in wood-based columns, using Engineering Village and Web of Science databases and PRISMA protocol to follow the procedures and ensure the quality of sampling. Reinforcement dispositions and types of assessments were identified so that the literature synthesis can contribute to identifying behavior models. Different methods of reinforcement sizing studied by the literature were synthesized and detailed as to their respective uses. A positive correlation between the reinforcement index and the increasing the load capacity of timber columns were discussed and statistically analyzed.

DOI: 10.15376/biores.20.1.2293-2303

Keywords: Structural recovery; Fiber-reinforced polymer; Mechanical performance; Column; Wooden construction; Structural reinforcement

Contact information: a: Department of Civil Engineering, Federal University of Rondônia, Porto Velho, Brazil; b: Department of Civil Engineering, Federal University of São Carlos, São Carlos, Brazil; c: Campus of Ariquemes, Federal Institute of Education, Science and Technology of Rondônia, Ariquemes, Brazil; d: Department of Engineering and Infrastructure, Federal Institute of Education, Science and Technology of Rondônia, Porto Velho, Brazil; e: Science and Engineering Institute, São Paulo State University, Itapeva, Brazil; f: São Carlos Engineering School, University of São Paulo, São Carlos, Brazil; Corresponding author: eng.pedrojardim@gmail.com

INTRODUCTION

Columns are used in timber construction typologies, which are indispensable to erect post-and-beam, post-and-lintel, and stick-frame systems. Wood is a bioresource, which explains why it is subject to defects, cracks, and deterioration that influence its durability, load capacity, and stability (Chen and Guo 2017; De Araujo 2023). Studies have reported visible degradations in wood columns (Qiao et al. 2016; Li et al. 2021), where significant interventions were essential to ensure the structural safety and in-service performance of the degraded elements. According to Valle et al. (2013), the use of fiber-reinforced polymers (FRP) is a viable alternative for structural reinforcement of wooden elements.

The use of FRP in reinforced concrete structures has been widely investigated (Amran et al. 2018), with the possibility of adopting specific codes for the use of this solution in structural projects. When consulting the literature, there has been little research into the use of FRP to reinforce wooden columns, which has an impact on the evolution of the reinforcement method, creating a barrier to its use by designers.

The exposition of the current state of the art on the reinforcement of timber-based columns with FRP can present valuable information for the development of new studies, as well as contribute to more assertive evaluations of these elements through the elaboration of a synthesis of the analytical dimensioning proposals of this reinforcement method. Therefore, this study aims to carry out a systematic literature review and a bibliometric analysis using statistics on the use of laminated FRP composites to reinforce timber columns, as to the contribution to improving their mechanical performances.

METHODOLOGY: SYSTEMATIC LITERATURE REVIEW

The systematized review was carried out up to mid-2024. Three stages were used: study planning, analysis of collected publications, and data extraction. In planning, research questions, databases for identifying articles and the search string were defined.

For a wider search, words with similar meaning to the adopted terms were listed. The search was carried out based on the titles, abstracts and keywords of the publications. Two scientific database tools were adopted: Engineering Village and Web of Science. The used string was: (“timber” OR “wood”) AND (“strengthening” OR “reinforcement”) AND (“column” OR “pillar”) AND (“fiber reinforced polymer” OR “frp”). Two research questions were formulated: What is the influence of the physical and mechanical properties of FRP and timber on the column performance? What mathematical methods are adopted in the literature for the design of FRP-reinforced timber columns? After the exclusion of duplicate records in databases, those articles that answered the raised questions and that could contribute to this review were selected. In the selection filter, all titles, keywords, and abstracts of the articles extracted from the searches were read to identify those that are not eligible. In the eligibility filter, introduction, and conclusion sections of filtered papers were read. In the inclusion filter, all eligible articles were read in full and classified according to their suitability for the sample. The articles approved in the three filters were used to extract the data of interest. Using VOSviewer, all keywords were compiled to generate a map of relationship among topics of sampled articles. The final stage of the sample definition process consisted of data extraction, where the studies were classified according to the quality criteria presented in Table 1, aiming to analyze the sample qualitatively regarding the sample’s capacity to answer the research questions.

Table 1. Criteria and Descriptions Analyzed in the Literature

BIBLIOMETRIC AND QUALITATIVE ANALYSIS OF SAMPLES

After sampling process, 15 articles were selected, corresponding to 22% of the records from two databases, excluding duplications. Figure 1 presents a PRISMA flowchart. Articles not included in the sample were excluded because they did not use FRP as a reinforcement element or because they did not analyze timber columns.

Fig. 1. PRISMA flowchart

Articles were obtained to recover other wooden elements like panels and beams, outside the scope of this review. Figure 2 shows the distribution of articles by year. The years 2015 and 2021 had more publications in the studied topic. Still, the small number of annual publications on the subject is verified.

Fig. 2. Number of publications according to year for systematic literature review (SLR)

Regarding quality criteria (Table 1), the number of articles per criterion is presented in Fig. 3. Physical and mechanical properties of the FRP (Q2) were met by 100% of the sample, followed by criterion Q1, with 80% on wood properties.

Fig. 3. Classification of studies according to quality criteria

The FRP was obtained from specialized manufacturers, having more production controls. This may explain the ease in obtaining and providing the values of the composite properties. Few studies performed a complete characterization. This may be explained by the difficulty in characterizing wood adequately, requiring several tests with distinct equipment to obtain different properties. Only longitudinal modulus of elasticity and compressive strength have been assessed. Criterion Q3 was met by 7 sampled articles (46.7%), indicating that there is a concern in proposing analytical design techniques for FRP-reinforced columns.

The map in Fig. 4 relates the keywords used together, showing the number of repetitions through the size of the word and the evolution of the terms adopted over the years. “Timber column”, “reinforcement”, and “composite reinforcement method” were the most frequently terms found. This indicates that the terms defined for the composition of the search string were suitable for this review, insofar as distinct terms adopted by the sample could emerge eventually. Although there were distinct groupings, keywords were similar, indicating axial compression tests, with variations in the type of fiber in the laminated composite and some indications of numerical simulations. The low number of articles (Fig. 1) and the similarity of the topics addressed (Fig. 4) indicate the importance of carrying out new studies that can elucidate possible topics not yet addressed.

Fig. 4. Map of relationship of keywords of sampled literature

ANALYSIS OF MECHANICAL PERFORMANCE OF REINFORCED COLUMNS

Using the data from sampled studies, it was possible to establish physical and geometric indexes of the reinforced columns with the mechanical performance found by the authors. For this, the volumetric reinforcement (V_R) was obtained using Eq. 1, and the ratio between the elasticity modulus of the FRP and the wood (E_FRP/E₀),

(1)

where A, V, L, t, p, and n denote, respectively, the cross-sectional area, volume, length, thickness, perimeter, and number of FRP layers. The consideration of V_R ratio to analyze the reinforcement in columns is considered adequate and is also used by Zhou et al. (2021), since it can be used both for columns with continuous reinforcement, covering the entire surface of the element, and for reinforcement with intervals, with FRP layers spaced apart. The reinforcement ratios of columns were related to the variation found for the load capacity (∆f) of structural element, as follows (Eq. 2).

(2)

Some properties, such as thickness, number of FRP layers or E₀, were not presented in the following studies: Chen et al. (2020), Dong et al. (2015a), Li et al. (2013), Najm et al. (2007), Xiong et al. (2016), and Zhu et al. (2013). Xiong and Su (2015) was not considered because it consisted of reinforcement in column joints by notch, using an intact column (without joint) as a reference model, and it was not possible to associate the results of the authors with the other studies. Even so, it was possible to obtain complete information on 42 columns tested in the laboratory. From these indexes analyzed, V_R ratio presented the highest levels of variation, with an average value of 5.05%, standard deviation of 8.52%, and coefficient of variation of 168.59%. This indicates that the literature has applied reinforcement with different proportions, which is considered positive because it allows evaluation of the impact of this variation on the reinforcement performance. Yang et al. (2021) adopted the highest V_R ratio, with 29.51%, in which a complete envelope of the external face of column was considered, as well as it was inserted the reinforcement in internal regions. The E_FRP/E₀ average ratio was 12.89%, with a standard deviation of 7.64% and coefficient of variation of 59.28%. The average indexes found can be considered as a starting point for reinforcement analyses in real timber columns.

The relation of reinforcement indexes (V_R and E_FRP/E₀) with ∆f is shown in Fig. 5. Since these indexes are related to different quantities (V_R = geometric and E_FRP/E₀ = physical), their product was also considered. A higher concentration of models with V_R less than 5% and E_FRP/E₀ less than 20% was found. For reasons of graphic scale, in which the ∆f data is grouped around V_R between 0 and 5% (Fig. 5a), this is a linear and unique behavior considering all individual results. Graphs did not show a clear pattern of dispersion, not even when considering the product of the reinforcement indexes.

Fig. 5. Dispersion between reinforcement indices V_R (a), E_FRP/E₀ (b) and V_R∙E_FRP/E₀ (c) with ∆f

It was possible to perform the correlation test by defining Spearman’s correlation coefficient (r_s) to evaluate the non-linear relationship between the V_R and E_FRP/E₀ and ∆f. These results are given in Table 2, where the bold values are those that presented correlation. Spearman’s correlation test showed that there is a weak positive correlation between the product of the V_R and E_FRP/E₀ ratios and the increase in the load capacity of fully FRP-clad columns. The p-value indicates that this correlation is significant (α<5%).

Table 2. Correlation of Reinforcement Indexes with the Increase in Resistance

It should be noted that Fig. 5a shows two trends in behavior: one for V_R < 3% and the other for V_R > 3%. Figure 6 shows a scatter plot for each group, accompanied by a regression model to predict ∆f. The group of pillars with V_R < 3% (Fig. 6a) showed no correlation between V_R and ∆f. On the other hand, the group with V_R > 3% (Fig. 6b) showed a linear relationship with ∆f, with a high Pearson correlation and a regression model with R²adj of 0.83. The pillars in Fig. 6b are made up of eight elements studied by Yang et al. (2021) and another three studied by Liu et al. (2022) and Wang et al. (2023).

Fig. 6. Dispersion between the V_R with increasing column resistance: V_R < 3% (a) and V_R > 3% (b)

ANALYTICAL METHODS FOR DESIGN

Articles with design methods were analyzed and their approaches are presented and compared in this chapter. In total, 8 articles presented analytical methods for predicting the mechanical performance of FRP-reinforced timber columns. Most of these articles were dedicated to presenting formulations for predicting the load capacity. Some studies (Li et al. 2013; Xiong and Su 2015; Xiong et al. 2016) took as a basis formulations developed for confined reinforced concrete columns, being a material with a greater volume of study than wood columns. Li et al. (2013) adopted the study by Mirmiran and Shahawy (1997) on the confinement provided by FRP composites in reinforced concrete-based columns to discuss the results found in the laboratory. After adapting the formulation for the materials adopted, Li et al. (2013) presented Eq. 3 to predict the additional compressive strength provided by the reinforcement,

(3)

where R and ε_FRP denote the radius of wooden column (circular section) and the FRP deformation, respectively. Although aforementioned authors used Eq. 3 to justify the results found, where the columns reinforced with FRP with a higher modulus of elasticity showed a greater gain in strength.

Xiong et al. (2016) and Xiong and Su (2015) also substantiated the adopted equation in a study on concrete columns. Xiong’s proposal adopts a confinement intensity coefficient (β) and a resistance reduction factor due to column cracks (γ). The adapted model is presented in Eq. 4, where b_FRP denotes the width of the FRP laminate and υ the Poisson’s ratio of timber.

(4)

Fig. 7. Diagram of reinforcement

In Xiong and Su (2015), the modification of the calculation model, originally proposed for concrete columns (Richart et al. 1929), was performed to predict the compressive strength of FRP-reinforced columns with joints (Fig. 7). Equation 5 presents the proposed modification,

(5)

where γ and β are, respectively, the column resistance reduction coefficient and the FRP restraint intensity coefficient, and r, b_f, and s denote, respectively, the column radius, the column reinforcement blade length and the spacing between the laminates (Fig. 8). The resistant force (P) of this connection was determined considering the calculation parameter presented in Eq. 6, resulting in Eq. 7.

(6)

(7)

Dong et al. (2015b) proposed a formulation for determining the maximum load of FRP-reinforced timber columns, presented in Eq. 8, considering the following hypotheses: perfectly concentric loading until failure, no relative sliding between FRP and the column, and failure due to shear cracks,

(8)

where α is a regression coefficient defined as 0.61 and b is a conversion coefficient equivalent to -0.05. The proposed model requires few properties, most of which are geometric, as is the case with the model by Xiong and Su (2015) and Xiong et al. (Xiong et al. 2016), facilitating its use when equipment for a full characterization of the wood is not available. When comparing the experimental results obtained with those estimated by Eq. 8, Dong et al. (2015b) found errors of up to 20%.

Liu et al. (2022) observed that the sizing method present in the Chinese code (SAC 2017) does not present adequate results for the strengthening of glued timber columns with loss of cross-section.

This inaccuracy is evinced by Zhou et al. (2021). Thus, Liu et al. (2022) presented a modification of the calculation method in Eq. 9,

(9)

where F_aB and F_a are the ultimate axial load capacity for hollow timber columns reinforced with BFRP and the ultimate axial load capacity of non-reinforced elements (kN), respectively; β_B denotes the adhesion coefficient of BFRP. Determined according to the Chinese code (SAC 2017), F_a is presented in Eq. 10 and β_B in Eq. 11.

(10)

(11)

From Eq. 10, φ is the axial stability coefficient (Eq. 12), A_n is the net cross-sectional area of the column, and f_c denotes the compressive strength of timber. From Eq. 11, ρ is the reinforcement ratio of the BFRP (A_FRP / A_timber), b is the length of the reinforcing blade, and d is the distance between the reinforcing blades, both relative to the longitudinal direction of the column, as shown in Fig. 7.

From Eq. 12, β, a_c, b_c, and c_c are correlation coefficients obtained from the Chinese code (SAC 2017), λ is the slenderness of the compressed element, and E_k and f_ck consist of the characteristic modulus of elasticity and compressive strength.

(12)

Fig. 8. Indication of the variables adopted for determining β_B

By adopting Eq. 9, Liu et al. (2022) estimated the resistance of the five reinforced columns studied, with an error ranging from -1.38% to 1.67%. The equation proposed by these authors is considered simple and is applied to columns with reinforcements at intervals. However, it depends on coefficients presented in the Chinese code (SAC 2017), which must be evaluated in order to be applied in other territories.

In the study conducted by Yang et al. (2021), where the reinforcement was considered in the internal region of the column, the proposed equation was based on the Tsai-Wu failure criterion, obtaining Eq. 13.

(13)

where F₁, F₁₁, F₁₂, F₂₂, F₂₃, F’₁, F’₁₂, and F’₂₂ consist of strength parameters calculated by the Tsai-Wu method, p_w is the restraining stress acting on the timber core (Eq. 14), A_l is the cross-sectional area of the internal reinforcement, and f_t3 is the transverse resistance of the internal FRP. L_c is the length of the wood core segment and m is the confinement coefficient.

(14)

To apply the method of Yang et al. (2021), a complete characterization of the wood and its reinforcement is necessary to determine the resistance parameters, which depend on the transverse and longitudinal tensile and compressive strengths of the wood, as well as the shear strength. Therefore, the method may be difficult to apply in situations where all this information is not available. Even so, the error found by the authors varied between -17.6% and 8.9%, being similar to the other models found in this review.

CONCLUSIONS

1. A weak positive correlation was found between the reinforcement ratios (product of the volumetric reinforcement rate and the E_FRP/E₀ ratio) and the increase in the load capacity of the fully FRP-clad columns.
2. The literature provides some calculation models for FRP-reinforced wooden columns with their accuracy varying by up to 20%, depending on the model.
3. In order to obtain results that allow conclusions to be drawn about the influence of reinforcement on the increase in compression load, parametric numerical simulations are needed to support a proposal for an equation that allows this increase in force to be estimated with greater precision.
4. Through bibliometric analyses, it was possible to identify gaps for further studies, including evaluation of the proposed models for estimating the load capacity of FRP-reinforced columns/pillars when adopting different types of fibers to reinforce the composite. Also needed are investigations into the use of hybrid FRPs in different applications, including degraded columns, with square cross sections, in joint regions, in glued wood-based elements, and others.

ACKNOWLEDGMENTS

The authors acknowledge the Departamento de Pesquisa, Inovação e Pós-Graduação of Instituto Federal de Rondônia, Campus Ariquemes and the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brazil (CAPES).

REFERENCES CITED

Amran, Y. H. M., Alyousef, R., Rashid, R. S. M., Alabduljabbar, H., and Hung, C.-C. (2018). “Properties and applications of FRP in strengthening RC structures: A review,” Structures 16(July), 208-238. DOI: 10.1016/j.istruc.2018.09.008

De Araujo, V. (2023). “Timber construction as a multiple valuable sustainable alternative: Main characteristics, challenge remarks and affirmative actions,” International Journal of Construction Management 23(8), 1334-1343. DOI: 10.1080/15623599.2021.1969742

Chen, Q., Chen, C., Yu, F., Shen, B., and Tan, C. (2020). “Experimental study on CFRP strengthening hollow wooden column and strengthening effect,” IOP Conference Series: Earth and Environmental Science 455(1), 012009. DOI: 10.1088/1755-1315/455/1/012009

Chen, Y., and Guo, W. (2017). “Nondestructive evaluation and reliability analysis for determining the mechanical properties of old wood of ancient timber structure,” BioResources 12(2), 2310-2325. DOI: 10.15376/biores.12.2.2310-2325

Dong, J. F., Jia, P., Yuan, S. C., and Wang, Q. Y. (2015a). “Compressive behaviours of square timber columns reinforced by partial wrapping of FRP sheets,” Materials Research Innovations 19(March), S1465-S1468. DOI: 10.1179/1432891715Z.0000000001593

Dong, J. F., Yuan, S. C., Wang, Q. Y., and Liang, W. (2015b). “Influence of fractured wood texture on structural behaviour of timber columns with fibre reinforced polymer reinforcement,” Materials Research Innovations 19(March), S5546-S5550. DOI: 10.1179/1432891714Z.0000000001149

Li, L., Yuan, S. L., Dong, J. F., and Wang, Q. Y. (2013). “An experimental study on the axial compressive behavior of timber columns strengthened by FRP sheets with different wrapping methods,” Applied Mechanics and Materials 351-352, 1419-1422. DOI: 10.4028/www.scientific.net/AMM.351-352.1419

Li, S., Song, T., Milani, G., Abruzzese, D., and Yuan, J. (2021). “An iterative rectification procedure analysis for historical timber frames: Application to a cultural heritage Chinese Pavilion,” Engineering Structures 227(August 2020), article 111415. DOI: 10.1016/j.engstruct.2020.111415

Liu, R., Wu, Z., Peng, Q., Zhang, Y., and Wang, J. (2022). “Experimental study on the axial compression performance of glued wood hollow cylinders reinforced with BFRP,” Sustainability 14(24), article 16827. DOI: 10.3390/su142416827

Mirmiran, A., and Shahawy, M. (1997). “Behavior of concrete columns confined by fiber composites,” Journal of Structural Engineering 123(5), 583-590. DOI: 10.1061/(ASCE)0733-9445(1997)123:5(583)

Najm, H., Secaras, J., and Balaguru, P. (2007). “Compression tests of circular timber column confined with carbon fibers using inorganic matrix,” Journal of Materials in Civil Engineering 19(2), 198-204. DOI: 10.1061/(ASCE)0899-1561(2007)19:2(198)

Qiao, G., Li, T., and Chen, Y. F. (2016). “Assessment and retrofitting solutions for an historical wooden pavilion in China,” Construction and Building Materials105, 435-447. DOI: 10.1016/j.conbuildmat.2015.12.107

Richart, F. E., Brandtzaeg, A., and Brown, R. L. (1929). The Failure of Plain and Spirally Bound Concrete in Compression, University of Illinois Engineering Experiment Station Bulletin, Urbana.

SAC. (2017). “GB 50005: Code for design of timber structures,” National Standard of the People’s Republic of China, Beijing.

Valle, Â. do, Pereira, N. B., and Nappi, S. C. B. (2013). “Recommendations for design in preserving historical buildings with wooden structures’ roofs,” Advanced Materials Research 778, 1072-1079. DOI: 10.4028/www.scientific.net/AMR.778.1072

Wang, H., Chun, Q., Zhang, C., Ma, S., Cao, G., and Dong, Q. (2023). “Experimental study on the compression behavior of long timber column strengthened with the novel hybrid fiber sheets,” Journal of Wood Science, Springer Nature Singapore, 69(1). DOI: 10.1186/s10086-023-02077-2

Xiong, X. Y., Lu, X. X., and Xue, R. J. (2016). “The study on timber pier columns strengthened with CFRP hoops in ancient architecture,” Advanced Composites Letters 25(6), 132-136. DOI: 10.1177/096369351602500602

Xiong, X. Y., and Su, Z. Y. (2015). “Experimental study and theoretical analysis of carbon fibre-reinforced polymer strengthening timber pier column,” Materials Research Innovations 19(March), S51246-S51254. DOI: 10.1179/1432891714Z.0000000001288

Yang, L., Li, X., Fang, H., Liu, W., Hong, J., Hui, D., and Gaff, M. (2021). “Compressive behaviour of wood-filled GFRP square columns with lattice-web reinforcements,” Construction and Building Materials 310. DOI: 10.1016/j.conbuildmat.2021.125129

Zhou, L., Liu, Q., Ma, S., and Han, X. (2021). “Eccentric compression behavior of long poplar columns externally reinforced by BFRP,” Journal of Wood Science 67(1), 1-17. DOI: 10.1186/s10086-020-01934-8

Zhu, Y. M., Yuan, S. C., Hou, M., and Wang, Q. Y. (2013). “Square short wood columns strengthened with FRP sheets under compressive load,” Applied Mechanics and Materials, 256-259(PART 1), 1008-1011. DOI: 10.4028/www.scientific.net/AMM.256-259.1008

Article submitted: December 17, 2024; Peer review completed: January 25, 2025; Revised version received: January 27, 2025; Accepted: January 29, 2025; Published: February 1, 2025.

DOI: 10.15376/biores.20.1.2293-2303