Abstract
Wood, as the oldest building material, provides some of the basic needs of human beings, including shelter and protection. Wood is used in exterior cladding, carrier systems, joinery, ceiling-floor coverings, windows, doors, and furniture production. When wooden material is exposed to external weather conditions, due to its hygroscopic structure, its physical and mechanical properties deteriorate from exposure to moisture, temperature, and biological organisms. The bending modulus of elasticity of Scots pine (Pinus sylvestris L.), oak (Quercus petraea L.), and chestnut (Castanea sativa M.) wood that was tannin-impregnated and heat-treated at 160 °C, was investigated using Taguchi L9 (33). The sequence was optimized. After heat treatment, the carrier elements were subjected to artificial climate conditions. In the optimization of the data obtained, it was understood that the highest impact factor was the tree type. In contrast, the climate on the elastic modulus was the lowest impact factor. In Taguchi analysis, a mathematical prediction model was created using actual and predicted data using the S/N ratio’s biggest-best equation. The R2 of the model can be predicted with an accuracy rate of 98.6%.
Download PDF
Full Article
Effect of Climate and Wood Type on Elastic Modulus of Heat-treated Wood and its Optimization by the Taguchi Method
Mehmet Gunes,a,* Cagatay Ersin,b and Mustafa Altunok c
Wood, as the oldest building material, provides some of the basic needs of human beings, including shelter and protection. Wood is used in exterior cladding, carrier systems, joinery, ceiling-floor coverings, windows, doors, and furniture production. When wooden material is exposed to external weather conditions, due to its hygroscopic structure, its physical and mechanical properties deteriorate from exposure to moisture, temperature, and biological organisms. The bending modulus of elasticity of Scots pine (Pinus sylvestris L.), oak (Quercus petraea L.), and chestnut (Castanea sativa M.) wood that was tannin-impregnated and heat-treated at 160 °C, was investigated using Taguchi L9 (33). The sequence was optimized. After heat treatment, the carrier elements were subjected to artificial climate conditions. In the optimization of the data obtained, it was understood that the highest impact factor was the tree type. In contrast, the climate on the elastic modulus was the lowest impact factor. In Taguchi analysis, a mathematical prediction model was created using actual and predicted data using the S/N ratio’s biggest-best equation. The R2 of the model can be predicted with an accuracy rate of 98.6%.
DOI: 10.15376/biores.19.2.3138-3148
Keywords: Wooden; Tannin impregnation; Modulus of elasticity in bending; Taguchi method; Optimization
Contact information: a: Design Department, Vocational School, Cankiri Karatekin University, Cankiri, Turkey; b: Electronics and Automation Department, Vocational School, Cankiri Karatekin University, Cankiri, Turkey; c: Wood Product Industrial Engineering, Faculty of Technology, Gazi University, Ankara, Turkey *Corresponding author: mehmetgunes@karatekin.edu.tr
Throughout history, humanity has effectively used natural, renewable, and sustainable materials, such as wood in many areas of life due to their advantages, such as aesthetics, high impact resistance, and low cost (Birinci et al. 2021; Yumrutaş et al. 2021). Because of wood’s abundance, its use and production are increasing (Imren 2022). Wood deteriorates when exposed to natural conditions without any protective treatment (De la Caba et al. 2007). The biological and hygroscopic structure of wooden materials makes such materials sensitive to various destruction factors and biotic factors (fungi, insects, termites, etc.). This essential factor limits wood’s use as a long-lasting building material. In particular, the effect of biotic factors on the mechanical properties of wood can negatively affect axial compression and bending resistance (Dimou et al. 2017). Additionally, wooden materials have an anisotropic structure. The anisotropic structure causes wood to behave differently under different types of loads, such as bending, tension, and compression. This can lead to various problems in engineering applications and structural designs (Zhao et al. 2016). Such challenges require the development of various methods for processing and preserving wooden materials. Physical methods, such as chemical impregnation treatments and heat treatment, can improve the mechanical properties of wood and make it more resistant to biotic factors (Li et al. 2022).
To optimize the experiments applied to the physical and mechanical properties of wooden materials, optimization of the effect of accelerated weathering conditions on wooden surfaces has been applied, using the Taguchi method (Kurt and Can 2021). The effect of artificial weather conditions on wood surface roughness (Imren 2022), optimization of wood laser cutting parameters (Nguyen and Dang 2023), determination of wood surface roughness in CNC milling (Budakci et al. 2013), improvement of heat-treated wood coating performance (Hazir 2021), and development of heat treatment parameters in wood materials (Workie and Tsegaw 2021).
When studies in the literature were examined, it was seen that the mechanical properties, modulus of elasticity, and bending resistance of wooden materials of different structures were calculated. The Taguchi method has been a preferred way to determine the factors affecting the change in bending strength and elastic modulus of the wood material, which was heat treated after tannin impregnation under artificial climate conditions, and to reduce the number of experiments (Taguchi 1993). In this study, the Taguchi method was used to examine heat treatment, tannin impregnation, and artificial climate conditions to determine which of the identified factors significantly impact the elastic modulus resistance of wood in bending. Test samples were obtained from 1/5 scale Scots pine, oak, and chestnut wooden beams. The process parameter was divided into three groups: 1. Untreated wooden control samples, 2. Samples were heat-treated in the oven at 160 oC for 2 hours, and 3. Wooden beam samples were heat-treated in a 10% acorn tannin solution for 24 hours.
EXPERIMENTAL
The wood materials used in the study were selected from Scots pine, oak, and chestnut wood used in the furniture and woodworking industry. The samples were supplied from the same enterprise in Ankara Siteler industrial zone using a random selection method. Care was taken to ensure there were no knots, fungi, or mold on the trees used in the experiments. Air-dry and fully dry density values of the tree species used are given in Table 1. Among the tree species used in the experiments, the air dry density of Scots pine (Pinus sylvestris L.) was 0.46 (g/cm3), of oak (Quercus petraea L.) was 0.84 (g/cm3), and of chestnut (Castanea sativa M.) was 0.45 (g/cm3). The trees supplied in lumber form were kept in the workshop environment until they became equilibrated with the air. Wood samples were positioned in the conditioning cabinet at 65±5% relative humidity (RH) and 20±2 °C to reach a constant weight, following the DIN EN 554 (2016) method. 3.2 cm x 6.4 cm x 88 cm method.
Tannin impregnation
Acorn tannin is an environmentally friendly and natural material obtained by extraction from acorns, an oak tree fruit abundant in forest regions. Tannin is most used in the leather industry (Merdan and Eyupoglu 2022). Tannin impregnation was used in the independent variable, the treatment factor. Some samples were determined as control samples, and bending strength tests were performed on wooden beams without any treatment. Some of the other samples were heat treated at 160 °C, and the rest were tannin impregnated before heat treatment. The samples with tannin impregnation were then subjected to heat treatment at 160 °C for 2 h. Acorn tannin was used to impregnate the samples. The test samples were kept in a plastic container for 24 h by dipping in 100 L of 10% acorn tannin solution (Fig. 1).
Fig. 1. Samples with impregnation application
For the impregnation process to be efficient, all samples were positioned so they were submerged in the solution and the container did not touch the ground. The samples, whose impregnation was completed, were ready for heat treatment.
Heat treatment application
After impregnation, each wood species was placed separately in the oven on the shelves shown in Fig. 2. The oven temperature was increased to 100 °C in 6 h. Then, the drying process was carried out by increasing it to 130 °C in 12 h. After the drying phase, the temperature was increased to 160 °C in 6 h, kept at this temperature for two h, and then the cooling phase was started. In the cooling process of the last heat treatment stage, the temperature was reduced to ambient temperature in 12 h. Hot steam was sprayed into the furnace at 3 atm pressure for 5 s every 60 s during heat treatment.
Fig. 2. Heat treatment oven
Conditioning
An artificial conditioning cabin was used to control the temperature and moisture content of the wooden material, which varies according to climatic conditions. The relative humidity and temperature in the air conditioning cabin were adjusted to mimic the summer, winter, and spring seasons. After the heat treatment, each tree species was placed in a separate air-conditioning cabinet, as shown in Fig. 3. The test samples were weighed every 12 h and kept in the air-conditioning cabinet at 20 °C/65% (spring) to 40 °C/35% (summer) to 10 °C/50% (winter) relative humidity, respectively, until they reached a constant weight for each climatic condition according to the DIN EN 325 (1993) standard. The samples that reached a constant weight were individually removed from the air-conditioning cabinet, and experiments were conducted.
Fig. 3. Universal climate test device
Flexural strength tests
Flexural strength tests were carried out on prepared Scots pine, oak, and chestnut beam samples with 32 x 64 x 880 mm dimensions by ISO 13061-3 (2014). A universal test device conforming to DIN EN 325 (1993) was used for bending strength measurement. In the experiments, the spacing of the fulcrum points was set as 832 mm (Fig. 4), the force at breakage (Pmax) was read from the midpoint of the part with a forward speed of 2 mm/min, and the bending resistance was calculated according to Eqs. 1 and 2.
Fig. 4. Universal climate test device
(1)
(2)
where Pmax is the maximum force applied at the moment of fracture (N), P is the load equal to the difference between the arithmetic averages of the lower and upper limits of the loading (N), ΔF is the force difference in the elastic region (N), l is the spacing between the support points (mm), b is the width of the test sample (mm), h is the height of the test sample (mm), and f is the deflection in the net bending area (mm).
Taguchi experimental design
In addition to their statistical significance, experimental design methods are complementary supportive methods that can be used in all research and development applications, increasing quality, minimizing costs, and thus strengthening the reliability of the outputs (Balki et al. 2016). The Taguchi method is widely used in engineering and industrial problems (Chamoli 2015; Kaya et al. 2020). Dr. Taguchi introduced a solution to increase efficiency in performing experiments and evaluating applications (Ross 1988). In addition to being an experimental design method, the Taguchi method is a valuable technique for high-quality system design by reducing the number of experiments (Taguchi 1993). To analyze the test outputs obtained in the Taguchi experimental design method, the loss function is converted into a statistical measure of performance known as the signal-to-noise ratio (S/N ratio). In this way, the deviation between the data is calculated.
According to its characteristic feature, this ratio is calculated with the objective functions (Masmiati and Sarhan 2015) given in Eq. 3:
(3)
The orthogonal array is used to create a custom design to determine the entire parameter space with a small number of experiments. The S/N ratio is used to analyze experimental results obtained from orthogonal array design. In the calculation of S/N ratios, “smaller is better”, “larger is better”, or “nominal is better” methods are used, depending on the characteristic type (Abdulaziz et al. 2020). Because bending resistance and elastic modulus resistance values were desired to be the highest among the dependent variables, S/N ratios were calculated using the more significant, better formula in Eq. 1.
Table 1. Factors and Values at Different Levels
Wood type, processing type, climate type factors, bending strength, and elastic modulus dependent variables were designed according to 9 experiments in the Taguchi L9 (33) orthogonal array in Table 2.
Statistical analysis
Statistical analysis was performed using Minitab 21 software (Minitab LLC, State College, PA, USA).
Table 2. Taguchi L9 (33) Orthogonal Experimental Design Matrix
According to the S/N response ratios obtained in Table 3, the factor that effects the elastic modulus the most is the wooden, followed by the treatment, and the factor that affects the least is the climate factor. According to S/N ratios, it was understood that the wood factor had the most impact on the elastic modulus. It was found that the effect of climatic conditions on the elastic modulus was minimal.
Table 3. S/N Response Table for Factors (Larger is Better)
An analysis of variance (ANOVA) was applied to these values in Table 4. Wood species provided the highest contribution to elastic modulus values, with 84.1%. When the factors affecting the elastic modulus were analyzed at the P < 0.05 confidence interval level, wood type was found to be the independent variable with the highest effect on the elastic modulus resistance with 84.1%. In contrast, it was determined that other factors, except wood, were insignificant. The contribution of process type was 12.9%, and the lowest contribution was made by climate type with 1.3%. From this, it is understood that the climate type factor has the most negligible impact on the elastic modulus.
Table 4. Elastic Modulus Variance Analysis
In Fig. 5, the effect of S/N ratios on the factors was calculated. The highest elastic modulus values were detected in oak wood. The highest values in the processing factor were obtained in control samples without heat treatment. In terms of climate, the highest values were found in the test samples conditioned at 10 °C/50%.
Fig. 5. Taguchi analysis modulus of elasticity versus wooden, treatment, and climate
Both bending strength and modulus of elasticity (MOE) values decreased in heat-treated samples compared to pine (Herrera et al. 2021), oak (Kubovský et al. 2020), and chestnut (Gunes and Altunok 2022) wood control samples. A more significant decrease in the elastic modulus was observed as the temperature and processing time increased (Hasanagic et al. 2023). It was determined that the elastic modulus values of sessile oak wood were higher than Scots pine and chestnut wood (Todorović et al. 2022). Elastic modulus values of wood exposed to artificial air conditioning conditions varied depending on the amount of moisture it contained. (Stadlmann et al. 2020). In line with previous studies, the elastic modulus values of oak wood were higher than those of Scots pine and chestnut wood among wood types and were parallel to the literature. The MOE values of treated wood decreased compared to control samples. There was a slight increase in tannin-impregnated samples compared to treated samples.
Fig. 6. Polynomial regression analysis: modulus of elastic versus predicted value
The mathematical model of the elastic modulus is shown in Fig. 6. When the estimated values limit, confidence interval and regression curve were examined, the estimated values (Modulus of elasticity = 3174 + 0.3383 Predict + 0.000033 Predict2) were predicted with a high accuracy of 98.6% by the specified mathematical model.
This study investigated the effects of heat treated with tannin impregnation and artificially conditioned trees on the elastic modulus values in bending. An experimental design was made on the elastic modulus values with the Taguchi L9(33) octagonal matrix, and a mathematical prediction model was created by optimizing the values.
- According to the S/N ratio and ANOVA analysis, it was determined that the independent variable that most affects the elastic modulus in bending was the tree type. The climate parameter had the lowest impact. Elastic modulus values of unheat-treated wooden beams were higher than heat-treated and tannin-impregnated samples.
- It has been understood that the Taguchi design method is the appropriate analysis method to determine the elastic modulus of wooden materials by performing several experiments. It can be said that this methodology provides an accurate and effective method when determining the effective parameters in the mechanical properties of wood.
- Using the created mathematical model, and R2 value of 98.6% was reached, indicating high accuracy in estimating the factors affecting the elastic modulus. In this way, both test costs and time can be saved by reducing the number of experiments and the number of experiment repetitions.
- This study will contribute to the woodworking industry in selecting the parameters that most affect the elastic modulus and improving the mechanical properties of wood. The effects of tannin and similar impregnation substances applied to wood material on the modulus of elasticity can be determined by reducing the number of tests.
The response surface method can be applied in future studies by changing the number of dependent variables and factors. A mathematical model can obtain the desired dependent variable values by optimizing the factors.
ACKNOWLEDGEMENTS
The Gazi University Scientific Research Projects Unit obtained the data used in this research within the scope of project number 07/2019-24.
REFERENCES CITED
Balki, M. K., Sayin, C., and Sarikaya, M. (2016). “Optimization of the operating parameters based on Taguchi method in an SI engine used pure gasoline, ethanol, and methanol,” Fuel 180, 630-637. DOI: 10.1016/j.fuel.2016.04.098
Budakci, M., Ilce, A. C., Gurleyen, T., and Utar, M. (2013). “Determination of the surface roughness of heat-treated wood materials planed by the cutters of a horizontal milling machine,” BioResources 8(3), 3189-3199. DOI: 10.15376/biores.8.3.3189-3199
Chamoli, S. (2015). “A Taguchi approach for optimization of flow and geometrical parameters in a rectangular channel roughened with V down perforated baffles,” Case Studies in Thermal Engineering 5, 59-69. DOI: 10.1016/j.csite.2015.01.001
De la Caba, K., Guerrero, P., DelRio, M., and Mondragon, I. (2007). “Weathering behaviour of wood-faced construction materials,” Construction and Building Materials 21(6), 1288-1294. DOI: 10.1016/j.conbuildmat.2006.02.008
Dimou, V., Kaziolas, D. N., Zygomalas, L., and Avtzis, N. (2017). “Influence of biotic factors on the mechanical properties of wood, taking into account the time of harvesting,” Wood Material Science & Engineering 12(3), 140-148. DOI: 10.1080/17480272.2015.1063004
DIN EN 325 (1993). “Wood-Based panels – Determination of dimensions of test pieces,” German Institute for Standardization, Germany.
DIN EN 554 (1994). “Sterilization of medical devices – Validation and routine sterilization control by moist heat,” German Institute for Standardization, Germany.
Gunes, M., and Altunok, M. (2022). “Investigation of mechanical behavior of pre-heat treated wooden beams in different climate conditions,” Gazi University Journal of Science Part C: Design and Technology 10(2), 300-313. DOI: 10.29109/gujsc.1030245
Hasanagic, R., Fathi, L., Sefidruh, M., Bahmani, M., and Humar, M. (2023). “Mechanical performance of heat-treated Norway spruce (Picea abies) woDINdod,” Bulletin of the Transilvania University of Brasov 16(65), 105-114.
Hazir, E. (2021). “Improvement of heat-treated wood coating performance using atmospheric plasma treatment and design of experiments method,” Polymers 13(9), article 1520. DOI: 10.3390/polym13091520
Herrera-Builes, J. F., Sepúlveda-Villarroel, V., Osorio, J. A., Salvo-Sepulveda, L., and Ananias, R. A. (2021). “Effect of thermal modification treatment on some physical and mechanical properties of Pinus oocarpa wood,” Forests 12(2), article 249. DOI: 10.3390/f12020249
Imren, E. (2022). “Effect of weathering conditions on wood surface roughness: Optimal parameters determined via Taguchi analysis,” BioResources 17(4), 6668-6678. DOI: 10.15376/biores.17.4.6668-6678
ISO 13061-3 (2014). “Physical and mechanical properties of wood – Test methods for small clear wood specimens – Part 3: Determination of ultimate strength in static bending,” International Organization for Standardization, Geneva, Switzerland.
Kaya, A., Aslan, M., Yilmaz, N. F., and Kurt, H. I. (2020). “Taguchi analysis of apparent densities of Al-Mg-Sic composites,” El-Cezeri 7(2), 773-780. DOI: 10.31202/ecjse.695249
Kubovský, I., Kačíková, D., and Kačík, F. (2020). “Structural changes of oak wood main components caused by thermal modification,” Polymers 12(2), article 485. DOI: 10.3390/polym12020485
Kurt, R., and Can, A. (2021). “Optimization of the effect of accelerated weathering conditions on wood surfaces via the Taguchi method,” BioResources 16(1), 1642-1653. DOI: 10.15376/biores.16.1.1624-1653
Li, W., Zhang, Z., Yang, K., Mei, C., Bulcke, J. V., and Acker, J. V. (2022). “Understanding the effect of combined thermal treatment and phenol–formaldehyde resin impregnation on the compressive stress of wood.” Wood Science and Technology 56(4), 1071-1086. DOI: 10.1007/s00226-022-01400-2
Masmiati, N., and Sarhan, A. A. (2015). “Optimizing cutting parameters in inclined end milling for minimum surface residual stress–Taguchi approach,” Measurement 60, 267-275. DOI: 10.1016/j.measurement.2014.10.002
Merdan, N., and Eyupoglu, S. (2022). “Investigation of the effect of different mordant agents on dyeing of cotton fabric with natural dye obtained from gallnut,” İstanbul Ticaret Üniversitesi Fen Bilimleri Dergisi 22(43), 33-43.
Nguyen, H., and Dang, T. (2023). “Optimization of process parameters for wood laser engraving based on Taguchi method approach,” MM Science Journal 2023(3), 6723-6729. DOI: 10.17973/MMSJ.2023_10_2023053
Ross, P. J. (1988). “Taguchi techniques for quality engineering: Loss function, orthogonal experiments, parameter and tolerance design,” Data and Information Technology, McGraw-Hill, New York, NY, USA.
Stadlmann, A., Pramreiter, M., Stingl, R., Kurzböck, C., Jost, T., and Müller, U. (2020). “Durability of wood exposed to alternating climate test and natural weathering,” Forests 11(9), article 953. DOI: 10.3390/f11090953
Taguchi, G. (1993). “Robust technology development,” Mechanical Engineering-CIME, 115(3), 60-63.
Todorović, N., Popović, Z., Milić, G., Veizović, M., and Popadić, R. (2022). “Quality evaluation of heat-treated sessile oak (Quercus petraea L.) wood by colour and FT-NIR spectroscopy,” Wood Material Science & Engineering 17(3), 202-209. DOI: 10.1080/17480272.2020.1847188
Workie, A. B., and Tsegaw, A. A. (2021). “Optimization of treatment parameters to enhance bending strength of bamboo,” in: Advances of Science and Technology: 8th EAI International Conference, ICAST 2020, Bahir Dar, Ethiopia, Vol.2(8), pp. 315-325.
Yumrutas, H. I., Birinci, E., Yorur, H., and Atahan, A. O. (2021). “An innovative approach on the renewable hybrid barrier: Combined use of wood and sand,” CERNE 27, 1-10.
Zhao, S., Zhao, J. X., and Han, G. Z. (2016). “Advances in the study of mechanical properties and constitutive law in the field of wood research,” IOP Conference Series: Materials Science and Engineering 137(1), article ID 012036.
Article submitted: February 16, 2024; Peer review completed: March 16, 2024; Revised version received and accepted: March 27, 2024; Published: April 2, 2024.
DOI: 10.15376/biores.19.2.3138-3148