To measure the stress wave parameters of standing trees, Shastkalateh forest, an educational and research area in Gorgan (northern part of Iran) was selected as the study site. Fourteen healthy standing trees of 20 years Paulownia (Paulownia fortune), were designated randomly into two diameter classes (25-31 cm and 32-38 cm) at the breast height. In this forest, Paulownia trees have been planted and grown under a controlled plan by the Gorgan University Department of Forestry. For this reason, all Paulownia trees have been identified with associated technical records, and older trees have greater diameters than the younger trees.

According to previous studies, the density of wood of these trees is 290 kg/m³, with negligible differences among them, nevertheless, the density of wood was measured in all fourteen trees via unique samples for each tree.

Stress-wave velocities were measured in transverse and longitudinal directions using Electronic Hammer IML® equipment, that included two probes, two electronic screws, and two electronic sensors (start and stop sensors), a portable two-channel digital analyser, and a handheld electronic hammer. In the longitudinal direction, sensors were inserted 50 cm above and below the breast height (130±50) (Figure 1a) and measurement was done in two northern side and southern side of the trees. To measure the stress-wave speed in the transverse direction, two probes were installed radially at breast height (Figure 1b, Figure 1c), and measurement were conducted in two directions, north-south and east-west.

Figure 1:
Schematic of tree measurement using stress wave NDT (a) longitudinal direction, (b) transverse direction and (c) real photo.

The measured standing trees were cut down and three logs in length of 100 cm were separated from each tree and defined as L1: 80-180 cm, L2: 180-280 cm, and L3: 280-380 cm. Measurement of logs was done nearly 2 weeks after trees cutting down and wave speed was evaluated in the middle of the logs in both the radial and longitudinal directions (Figure 2). Thereafter, the MOE_d was calculated for standing trees and logs according to Equation 1 .

(1)

where MOE_d: dynamic modulus of elasticity (Pa), ρ: wood density (kg/m³), and V: stress wave velocity (m/s).

For greater confidence, the density of the samples was measured which was equal to 290 kg/m³.

Figure 2:
(a) Schematic and (b) real photo of log measurement using stress wave NDT.

The logs were then cut into small specimens with dimensions of 5×5×76 cm, followed by conditioning at 12% MC according to ASTM-D-143-94 (2007) standard. Several specimens were selected from each log in 4 geographical directions. The clear wood was selected with tangential patterns without defects and spiral grain. Finally, three-point bending tests were carried out on the samples to measure the modulus of elasticity (MOEs) using Equation (2):

(2)

where _^MOEs: static modulus of elasticity, L: span (m), _^Ppl: proportional limit load (N), ∆_pl: proportional limit displacement (mm), b: width of sample (m), and h: thickness of sample (m).

Table 1 shows the stress-wave velocities obtained from all standing trees. In general, the stress-wave speed in the longitudinal direction (2304 m/s) is greater than that in the transverse direction (1079 m/s), which might be attributed to the parallelism of the longitudinal wave with the fiber axis. These results are in line with the results of (Madhoushi and Daneshvar 2016) which showed that the particle oscillations are parallel to the direction of longitudinal wave propagation, whereas they are perpendicular to each other in transverse or shear waves.

Moreover, the speeds of both stress waves (longitudinal and transverse) in logs are greater than those in standing Paulownia trees (Figure 3, Table 1), which is consistent with previous report by (Madhoushi and Daneshvar 2016) and in spite of the other species such as the Scots pine (Pinus sylvestris L.), reported by (Auty and Achim 2008). This increase might be attributed to the moisture amount of wood in trees, which is naturally higher than that in logs. Previous findings reported that humidity is one cause of reduced wave speed in trees (Brashaw et al. 2004).

Moreover, the speed of stress wave in both longitudinal and transverse directions varies slightly from bottom to top of stem. It can be observed that the longitudinal stress wave increases at upper part of stem, however, the transversal stress wave after an increase in L1 decreases at upper parts. These results are consistent with previous results reported by (Ishiguri et al. 2013, Madhoushi and Daneshvar 2016), who demonstrated that speed of stress wave depends on the location of the stem. The reason for this variation in a transverse direction might be related to lower amount of wood materials at the upper parts of the stem.

Figure 3:
Stress wave velocity in trees and logs of Paulownia wood.

Table 1:
Stress wave velocity (m/s) in standing trees and logs in longitudinal and transverse directions.

L1, L2, L3: Stem height, L1: 80-180 cm, L2: 180-280 cm, and L3: 280-380 cm.

The results presented in Table 2 show that, in tree, the average amount of MOE_d in the longitudinal direction (1547 MPa) is higher than that in the transverse direction (359 MPa). Moreover, its magnitude in logs at the first height (S1) is significantly higher than that in standing trees, and its gradual increase can be seen at upper part of the stem. This trend can also be seen in transversal stress wave. The reason for this difference between tree and logs may be related to lower moisture content of logs (Wang et al. 2004), which in turns lead to higher speed of stress wave and MOE_d in logs compared to tree.

Table 2:
Average dynamic modulus of elasticity (MPa) in standing trees and three heights of stem in Paulownia.

L1, L2, L3: Stem height, L1: 80-180 cm, L2: 180-280 cm, and L3: 280-380 cm.

The mean value of the stress-wave speed is higher in logs of larger diameter (Figure 4) and this difference was statistically significant according to student's t-test. Thus, the MOE_d for diameter class 32-38 cm is more than that for diameter class 25-31 cm (Figure 5). This difference was also statistically significant. This might be due to large differences (wood amount) between two classes, and it is useful for practical purposes. These results are in line with those of (Wang et al. 2003, Wang et al. 2007), who found that the stress-wave speed increases in standing trees and logs with increase in diameter.

Figure 4:
Effect of log diameter on stress wave velocity.

Figure 5:
Effect of tree and log diameter on dynamic modulus of elasticity.

The results show that the mean value of modulus of elasticity of Paulownia wood is 5596 MPa (Figure 6). It shows that the strength of Paulownia is low and it can be considered as a weak wood. However, it is a fast-growing tree. It can be found that the modulus of elasticity increases slightly along the height of the stem. The reason for this variation in a transverse direction might be related to lower amount of wood materials at the upper parts of the stem. It might also may be affected by the diameter of the trees (Figure 7) because trees with larger diameters (32-38 cm) possess slightly higher MOE_s than trees with smaller diameters (25-31 cm). This finding shows that Paulownia trees planted in Iran require more time to grow to form a mature wood, is spite of poplar wood (Madhoushi and Daneshvar 2016). Generally, variation in the modulus of elasticity from bottom to top of stem depends on species, and in some species, there are similar trend, such as pine (Antony et al. 2011) and poplar (Madhoushi and Daneshvar 2016).

Figure 6:
Modulus of elsacticity in sawn wood of Paulownia.

Figure 7:
Effect of tree diameter on static modulus of elasticity.

Table 3 shows the degree of correlation between MOE_d in standing tree (as independent variable) and MOE_d in log at three levels (as dependent variable) in Paulownia. It can be found that the correlation between these two parameters is considerably strong only in longitudinal direction and in L1 (r=0,69) (Figure 8), and in transverse direction, all correlations are weak. It shows that NDT stress wave in transverse direction cannot be useful for prediction of properties. Strong correlation between these two dynamical parameters, which has also been reported for other species (Wang et al. 2000), can be very important from practical point of view, such as forest management and control of volume of wood production in trees.

Table 4 shows the degree of correlation between MOE_s (as dependent variable) of lumber and MOE_d in longitudinal direction (as independent variable) in standing tree of Paulownia. In addition, it can be observed that the general correlation between MOE_s of lumber and MOE_d in trees is considerably strong (r= 0,68) (Table 5). These results are consistent with the results of (Wang et al. 2004), who found that there is a good relationship between the MOE_d (in standing trees and logs) and MOE_s of lumber. This finding could be very useful for predicting MOEs by measuring MOE_d in the longitudinal direction in the standing tree.

It should be noted that the correlation between the MOE_d in tree (in transverse direction) and the MOEs was weak for all heights (Table 6). Therefore, this parameter cannot be useful for prediction of strength of lumber in Paulownia.

Figure 8:
Relationship between MOE_d in tree and log (L1) in Paulownia.

Table 3:
Correlation between MOE_d in standing tree and MOE_d in log at three levels in Paulownia.

L1, L2, L3: Stem height, L1: 80-180 cm, L2: 180-280 cm, and L3: 280-380 cm.

Table 4:
Regression equations between MOE_d in longitudinal direction and MOE_s in Paulownia trough stem.

L1, L2, L3: Stem height, L1: 80-180 cm, L2: 180-280 cm, and L3: 280-380 cm.

Table 5:
General relationship between MOE_d in longitudinal direction in tress and MOEs of lumber in Paulownia.

Table 6:
Regression equations between MOE_d in transverse direction and MOE_s in Paulownia.

^♠Corresponding author: madhoushi@gau.ac.ir