Printer Friendly

Annual ring orientation effect and slope of grain in hemlock timber drying.


This research was designed to evaluate the effect of annual ring orientation and slope of grain on the drying characteristic of 115-mm-square Pacific Coast hemlock timbers in a heat-and-vent kiln in an effort to reduce degradation and shorten drying time. The results suggested that the horizontal annual ring orientation pile arrangement shortened the drying time by 11.2 percent (2 days shorter) whereas the vertical one lengthened the drying time by 38 percent (7 days longer), compared to the control. Both vertical and horizontal orientations resulted in higher yield by at least 4.1 percent compared to the control. In each run, twist, bow, and diamonding changed significantly after the timbers were kiln dried when compared to their measured green deformations. After drying, the vertical ring orientation showed significantly greater bow change and shrinkage compared to the control and to other runs. The measured apparent slope of grain occurrence on the 800 specimens dried was not important as expected, resulting in a mean value of 1.2 degrees with a standard deviation of 0.34 degrees. Slope of grain effects on drying rates and shape distortions were inconclusive, probably due to the low value. Twist, bow, and diamonding after planing were significantly reduced for each run compared to their measured values before planing, but there were still 8.5 percent of timbers with bow and 7.8 percent of timbers with twist more than 3 mm.


Many wood attributes have been shown in the past to affect timber quality during, and in some instances, after kiln-drying. Among them, slope of grain (SOG), defined as the angle between the direction of the wood grain and the main axis of a lumber piece that could either be a consequence of the sawing process or a natural occurrence in trees or logs (spiral grain), can cause shape deformations. SOG in lumber reduces strength and could promote twist as the moisture content (MC) varies. In air-dried lumber cut from plantation conifers, twist increases with SOG and decreases with the distance of the board from the pith (Balodis 1972). Twist is caused by spiral grain in combination with anisotropic shrinkage and is most pronounced in small dimension lumber sawn near the pith for fast-grown Norway spruce and Sitka spruce according to Danborg (1994). It was also indicated that little was gained by grading lumber based on measuring SOG and that more benefit would be gained through a rough grading based on annual ring width, lumber dimension, and its position in relation to the pith. Typical average juvenile heartwood spiral grain is from 5 to 7 degrees; whereas for the wood near the bark it is only 2 degrees in radiata pine. The juvenile properties contributing to warp in the processing of young radiata pine are mainly spiral grain and longitudinal shrinkage (Haslett et al. 1991). Foresberg and Warensjo (2001) pointed out that SOG on the log surface accounted for a larger part of twist than SOG on lumber for Norway spruce. SOG has been proven to cause shape distortions during drying of lumber, but no work has been done on how it affects grade recovery and drying deformations in hem-fir baby squares.

Annual ring orientation (ARO) refers to the growth ring orientation relative to the main drying surfaces as seen from one of the timber's ends when in a kiln stack. Lumber sawing patterns affect the drying rate due to the inherent wood permeability differences between the radial and tangential directions. The ratio of radial over tangential specific permeability is 2 to 2.5 over 1 and this can considerably affect the drying rate during the constant rate and first falling rate periods. Past research has shown that in drying of radiata pine lumber, the quartersawn pieces need longer drying time than the flatsawn pieces. The drying time difference was 10 to 15 percent of the total drying time for conventional temperature drying, but less significant for high-temperature drying (Pang 2002). The results from McCurdy and Keey (2002) revealed that towards the end of high-temperature drying of radiata pine, the MC profile is flatter with quartersawn boards than with flatsawn material, but diminished permeability of the latewood bands reduces the drying rate compared with flatsawn lumber under the same drying conditions. The ARO arrangement in a stack of timbers has been used in some custom-drying operations with improved grade recovery (Avramidis 2001), but no systematic and quantifiable research has been done on this issue.


Materials and methods

Freshly cut 115 mm square and 3.05 m long Pacific Coast hemlock (hem-fir) green timbers were obtained from a local mill of standard and better grade. The green timber population of 800 pieces was divided into 2 groups based on the following criteria: for the ARO tests, the wood had at least 80 percent annual rings on the end of each piece connected to two opposite sides and no pith present; for the SOG tests, the specimens were picked by sorting 320 pieces and then choosing 160 pieces above the calculated median for the high-SOG group and the rest were used in the low-SOG group. The green timbers were cut to 2.44-m-long kiln specimens. One section (cookie) was cut from each end next to the trimmed end, for basic density and initial MC measurement (Fig. 1). The cut cookies were immediately weighted with an electronic balance to 0.01 g, then their volume was measured by water dipping, and finally they were ovendried at 103[+ or -]2[degrees]C to constant weight. The measurements were used for the determination of their green MC and basic density (ovendry weight divided by green volume). The mean value of the two cookies per specimen was used for each timber. Digital calipers were used to measure the green thickness and width at mid-length to 0.05 mm. The green weight of each timber was also measured with a digital scale. Shape distortions of each green timber were then evaluated as explained below (as in the post-drying).

Measurements of SOG were carried out with a shop-built scratch gauge with pivoting handle for scribing a light groove parallel to the grain. A MITE-R-GAGE adjustable Lexan protractor was used for measuring the angle between the timber edge and the groove, measurable to 0.5 degrees. For each piece of timber, the SOG was measured at three points: front-quarter-end, middle-point, and rear-quarter-end. The surface groove made by the scratch gauge had a minimum length of 300 mm so that an average SOG, instead of a local one, could be measured.

A 7-[m.sup.3] steam-heated and track-loaded dry kiln was used in this study. Five drying runs (Table 1) were carried out, each comprised of 160 specimens, where V-ARO means annual rings of timber are vertical to the main drying surfaces in the kiln and Ho-ARO means the rings are horizontal. Each stickered green timber package was comprised of 8 pieces high and 20 pieces wide and the timbers were stacked tight from edge to edge. The kiln cart sits on a scale that allows for continuous charge weight monitoring and thus its MC. The 10-step drying schedule used is shown in Table 2. Air velocity was set to 3 m/sec. through the timber stack measured with a flow meter; flow direction was reversed every 12 hours. The target MC was set at 15 percent.

Upon completion of each drying run, the timber load was cooled down in the kiln for 12 hours with the doors closed. Then all specimens were pulled out, reweighted, and their thickness and width were re-measured. A Delmhorst resistance moisture meter was used to measure the dry MC of each timber at 25-mm (shell) and 50-mm (core) depths.

Shape distortions after drying (twist, bow, and diamonding) were measured on a shop-built aluminum table. It consists of a "U" shape aluminum base that is clamped, in an upside down position, onto two support stands with leveling feet. The flatness of the surface of the base was ground to 0.25 mm. An aluminum fence was lapped, shimmed, and mechanically clamped at 90[degrees] to the base. Its straightness was adjusted to 0.25 mm with a piece of 19-mm medium density fiberboard (MDF) sawn on a very accurate sliding table saw and a feeler gauge. The straightness of the MDF edge was checked against a ground steel machine surface to be within 0.05 mm. The measuring table was fixed to the same location in the lab where the table was adjusted for straightness and flatness. The angle between the fence and the base was adjusted to 90[+ or -]0.5 degrees.

In conjunction with the measuring table, two special shop-built digital dial gauges were used to measure bow, twist, and diamonding. The first two deformations were measured with a Mitutoyo Model ID-C1012EB Digital Dial Gauge attached at 90 degrees to a flat aluminum reference plate. The resolution was 0.01 mm and the measurement accuracy was better than [+ or -]0.5 mm in the hands of a skillful operator. Measurement of diamonding was carried out with a shop-built instrument consisting of a Mitutoyo Model ID-S1012EB Digital Gauge attached to a precision steel machine square. The resolution is 0.01 mm and the measurement accuracy was better than 0.25 mm.

Once all five runs and the specimen post-drying evaluations were completed, the wood was shipped to a local mill where it was planed and then graded by a professional grader. The grader followed the grading standards for the current Japanese markets. Following planing, the dimensions and shape deformations of each timber were once more evaluated as described previously.

Results and discussion

The measured basic density/standard deviation for the control, V-ARO, Ho-ARO, H-SOG-Ho-ARO, and L-SOG-Ho-ARO was 420/50, 390/40, 400/40, 400/50, and 400/50 kg/[m.sup.3], respectively. The initial MC of the five runs is listed in Table 3. Although the values for the first three runs appear different, t-test analysis shows that their difference is statistically insignificant. The t-test also showed no significant difference between the mean MC values for H-SOG-Ho-ARO and L-SOG-Ho-ARO.

The difference between minimum and maximum MC is at least 105 percent with a standard deviation of more than 20 percent in each run. This is not unusual for hem-fir and is a result of wet pockets and green moisture differences between hemlock and fir that make the uniform drying of baby squares difficult without bearing high operational costs and degradation. In each run, the same target final MC of 15 percent was set in the schedule to command the kiln to go to the conditioning step. The final MC values (KD rough) based on the recording of the scale under the kiln for the five runs are listed in Table 4. Actual shell-and-core MCs and their differences are also listed. The disparity between the minimum and maximum MC (KD rough) in each run was quite high although a long and mild (at the early drying stages) schedule was used. Still, this is not unusual, but merely a reflection of the difficulty in drying coastal hem-fir baby squares because of the wide spread of initial MC and mixed density. The presence of wet pockets (very high MC areas within the heartwood) is another important factor that makes drying difficult. In Table 4, the negative difference between shell-and-core MC indicated the existence of wet pockets.

The measured SOG and standard deviation (in parentheses) of the control, V-ARO, Ho-ARO, H-SOG-Ho-ARO, and L-SOG-Ho-ARO were 1.35 (0.84), 1.08 (0.68), 1.57 (0.72), 1.52 (1.08), and 0.31 (0.34), respectively. T-test analysis for the ARO and SOG groups showed that all runs were statistically different at a probability value of at least 0.056, as compared to the control run. When the total five-run population was examined, the average SOG of the 800 pieces was 1.2[degrees] with a range from 0.0[degrees] to 7.83[degrees] and very few pieces (less than 5%) with SOG of more than 5[degrees]. Mostly, SOG ranged from 0 to 2.5[degrees].

The drying times from the scale reading of the kiln for each run are listed in Table 5. The drying time differences compared to the control run are also listed. The drying time change as compared to the control run was 38.6 percent and -11.4 percent for V-ARO and Ho-ARO runs, respectively. This means that the Ho-ARO pile arrangement shortenened the drying time by 11.4 percent (2 days shorter); whereas the V-ARO lengthened the drying time by 38 percent (7 days longer). This is the result of the higher permeability in the radial direction (Ho-ARO case) where the presence of ray-cells would assist in water transfer and evaporation during the constant rate and possibly the first falling rate periods. Radial permeability is about 2 to 2.5 times greater that the tangential one (Siau 1995) and therefore placing timbers in Ho-ARO arrangement will increase the drying rate and consequently decrease the total drying time. Furthermore, the presence of rays could also accelerate the diffusion of water molecules since the lumen diffusion coefficients are much higher than the cell wall ones. This phenomenon will become prominent partially in the first falling rate period and completely in the second falling rate period.

Figure 2 shows the normalized drying curves (left) and normalized drying rates (right) for the two ARO runs with the control. The normalization was done by using the fractional MC that is defined as the ratio of current over initial MC. Figure 2 shows the longer drying time for the V-ARO group and the shorter time for the Ho-ARO group, as compared to the control. It can be observed that the Ho-ARO run dried faster from the 3rd day to the 8th day and that the V-ARO run dried slower in this period of time, as compared to their control run. After the 8th day, this drying rate difference diminished. Before the 3rd day, the drying rate difference was not obvious because the drying schedule was formulated so there was almost no drying for the first two steps in order to prevent surface checking, and also to heat the timbers by keeping them "moist." The theory behind this strategy is to reduce internal moisture and temperature stresses and assist in "opening-up" the structure of wood, thus resulting in faster drying and lower degrade. Alternatively, the Ho-ARO run dried faster compared to the control run between 0.4 and 0.85 of fractional MC (Fig. 2, right). In this range, the V-ARO run dried slower than the control.

The average transverse linear shrink-ages for the control, V-ARO, Ho-ARO, H-SOG-Ho-ARO, and L-SOG-Ho-ARO were 3.02, 3.48, 2.51, 3.04, and 2.76 percent, respectively; the width/thickness shrinkage values for the five runs were 2.86/3.17, 3.97/2.99, 2.09/2.92, 2.43/3.64, and 2.26/3.26, respectively. They ranged from 2.51 to 3.48 percent with a mean value of 2.96 percent. The t-test results showed that any pair of the means is significant, except for the comparison between the control and H-SOG-Ho-ARO runs. Shrinkage for the V-ARO run was significantly greater than its control run and all other runs. An interesting comparison can be made here. When the width/thickness data are compared between the Ho-ARO and V-ARO configurations, we can see the evident difference between radial and tangential width shrinkage; radial was 3.97 percent while tangential was 2.09 percent. However, when the thickness shrinkage between the two groups is compared, this difference is no longer evident. This could be the result of the vertical force exerted by the weight of the timbers, which caused an override effect. This is a speculation that requires further investigation. Also, the width/thickness shrinkage ratio is in the reverse direction of the radial/tangential or tangential/radial anisotropy for the V-ARO and Ho-ARO configurations. One explanation for the Ho-ARO configuration could be that the tangential shrinkage is decreased following a permanent creep deformation taking place at the start of drying because of high tensile stresses at the main drying surface (tangential direction). This phenomenon is amplified by the steep MC gradients in the thick timber drying as in this study. In addition, this reverse trend is due to the fact that the wood is weaker in the tangential direction. For the V-ARO configuration, it is hard to explain. This could result from the creep deformation difference between the tangential and radial directions.


The measured shape distortion data for the five runs are listed in Table 6. In each run, twist, bow, and diamonding changed significantly after the timbers were dried as compared to their measured green deformations. After drying, the V-ARO run showed a significantly greater change in bow as compared to the control run and all other runs. The Ho-ARO run exhibited significantly less change in twist as compared to its control run. Both Ho-ARO and V-ARO runs exhibited significantly less change in diamonding as compared to their control run since the former two had their annual ring orientation parallel to the sides of each timber. The initial shape deformations measured also reveal that the green bow values were quite close to the maximum of 2.44 mm allowed by the Japanese pre-cutters. This value is calculated as 0.1 percent of the timber's length (2.44 m). Considering that Japanese carpenters will not allow bow more than a "bu" ([approximately equal to] 3 mm), more effort should be made by the industry to produce baby squares with less green deformation.

A mill professional grader evaluated the wood specimens after planing. There were seven grade categories used and the top three grades were collectively treated as Zairai (high quality wood for traditional Japanese house construction) grade. A cut-off limit of 24 percent MC was used for the wet grade and the wets were only assigned and labeled if they were expected to be of 1JST & grade after re-drying. All grading results are listed in Table 7. The grading outcome showed that for the top three grade yield, the Ho-ARO run was 7.2 percent higher compared to the control run, and that the V-ARO run was 7.9 percent higher compared to the control.

Since both H-SOG-Ho-ARO and L-SOG-Ho-ARO runs used the horizontal ARO pile arrangement, their drying time and yield were also compared to the control. Because these two runs had significantly greater initial MC than the other three runs, their drying times were adjusted and listed in Table 5 from 58.4 percent MC to the target 15 percent final MC so that they could be compared to the control. In this case, H-SOG-Ho-ARO and L-SOG-Ho-ARO exhibited shorter drying times by 14.1 and 8.2 percent, respectively, when compared to the control. Again, the horizontal ARO contributed to this result as explained before. In addition to the change in the drying time, the top three grade yield for the H-SOG-Ho-ARO and L-SOG-Ho-ARO runs was higher than their control run by at least 4.1 percent (Table 7), regardless of the SOG difference. The L-SOG-Ho-ARO run had 10 pieces and other runs had from 1 to 3 timber pieces of the Clear grade. The yield of the top three grades in the H-SOG-Ho-ARO run was lower than the L-SOG-Ho-ARO run by 2.8 percent. For all fives runs, no timber was assigned into reject and wet grades, indicating that the drying process was appropriate considering both the drying loss and timber moisture content uniformity.

Figure 3 shows that the H-SOG-Ho-ARO and L-SOG-Ho-ARO runs demonstrated almost identical drying curves (left) and similar drying rates (right). This indicated that this particular SOG had no effect on drying rate as expected, probably because it was not high enough to reveal any effect. Similarly to the two ARO runs, the two SOG runs with the horizontal pile arrangement dried faster for the first 12 days (Fig. 3, left) and their drying rates were higher, between 0.53 and 0.95 of fractional MC (Fig. 3, right), compared to the control run.

After drying, the twist change in the H-SOG-Ho-ARO run was not significantly different compared to the control run and there was no difference between the H-SOG-Ho-ARO and L-SOG-Ho-ARO runs. Although the V-ARO run had a SOG of 1.08 degrees, it exhibited the significantly greatest bow change after drying, compared to the other four runs. The lack of evidence regarding the effect of SOG on timber deformations is probably due to the low slopes found in this study. It is "unfortunate" that we could not find high SOG timbers because this part of the study is begging for further evaluation. Diamonding change after drying in the L-SOG-Ho-ARO run was significantly greater as compared to its control run and to the H-SOG-Ho-ARO run and to the other two runs. Between the H-SOG-Ho-ARO and L-SOG-Ho-ARO runs, only the diamonding change was significant. The Ho-ARO run had a SOG of 1.57 degrees, which is almost identical to the SOG of 1.52 degrees in the H-SOG-Ho-ARO run, but the changes in the twist and diamonding after drying were significantly higher with the H-SOG-Ho-ARO run than with the Ho-ARO run. From these mean comparisons, it seems that twist and bow changes from drying were independent of SOG in this low angle study. It is possible that with higher slope specimens we would have seen more noticeable effects.




In order to further evaluate the SOG effect based on its distribution within the H-SOG population, we attempted to look at individual deformation values rather than the means. Figures 4, 5, and 6 show shape distortions plotted against measured SOG values for specimens with at least a 2-degree angle. It can be seen that the [r.sup.2] values of linear regression were very low for twist, bow, and diamonding, thus indicating no apparent effect of SOG on these three deformations. However, the twist difference between the green timbers and kiln-dried (rough) ones showed a different scattering pattern within the top four timber rows (Fig. 7) compared to the bottom four rows (Fig. 8) of the kiln stack in the H-SOG-Ho-ARO run. This result demonstrated an effective way to control twist in drying with: dead weights on the top of the kiln load. Figure 8 shows even more pieces with reduced twist after the timbers were dried (negative values) compared to the measured green twist, indicating twist reversal in drying. This can happen if the timber SOG is oriented along the thickness instead of along the width direction of the timber.

Shape deformations after drying and planing are listed in Table 8. The t-test results indicated that twist, bow, and diamonding after planing were significantly reduced for each run. However, there were still 8.5 percent of timbers with bow more than 3 mm and 7.8 percent with twist more than 3 mm. Over all 800 pieces from the 5 runs, bow in utility grade appeared to be higher than that in the other four grades, indicating that bow reduction could be an effective way to lower degrade.


The following conclusions can be drawn in light of this investigation:

1. The horizontal annual ring orientation pile arrangement shortened the drying time by 11.2 percent (2 days shorter), compared to the control (random) run; whereas the vertical annual ring orientation pile arrangement lengthened the drying time by 38 percent (7 days longer) compared to the control run.




2. Both vertical and horizontal annual ring orientation pile arrangements resulted in higher yield by at least 4.1 percent compared to the control run.

3. In each run, twist, bow, and diamonding changed significantly after the timbers were kiln dried when compared to their measured green deformations.

4. The vertical annual ring orientation pile arrangement showed significantly greater bow and shrinkage compared to the control run and to the two slope of grain runs.

5. The measured apparent slope of grain occurrence on hem-fir baby squares was not important as expected, resulting in a mean value of 1.2 degrees with a standard deviation of 0.34 degrees for all 800 baby squares tested in this project.

6. Slope of grain effects on drying rates and shape distortions were inconclusive due to the low slope of grain levels that occurred in this study.

7. Timbers with high slope of grain resulted in a reduced twist difference on the bottom four rows in the kiln stack as compared to the timbers on the top four rows.

8. Twist, bow, and diamonding after planing were significantly reduced for each run compared to their measured values before planing, but there were still 8.5 percent of timbers with bow more than 3 mm and 7.8 percent of timbers with twist more than 3 mm.
Table 1. -- Details of the five drying runs.

Run Specifics

Control (run 1) Control--mixed SOG and mixed ARO
V-ARO (run 2) Mixed SOG and vertical ARO
Ho-ARO (run 3) Mixed SOG and horizontal ARO
H-SOG-Ho-ARO (run 4) High SOG and horizontal ARO
L-SOG-Ho-ARO (run 5) Low SOG and horizontal ARO

Table 2. -- Drying schedule used in the five runs. (a)

Step Time [T.sub.db] [T.sub.wb] [M.sub.e] Fan reversal
 (hr.) ([degrees]C) (%) (hr.)

 1 6 49 49 25 3
 2 24 52 51 21.5 12
 3 24 55 53 17.5 12
 4 24 58 55 15.6 12
 5 24 62 57 12.6 12
 6 24 66 59 10.6 12
 7 24 70 61 8.6 12
 8 24 74 63 7.8 12
 9 ... 78 65 7 12
10 12 72 69 15.1 6

(a) [T.sub.db] = dry-bulb temperature; [T.sub.wb] = wet-bulb
temperature; [M.sub.c] = equilibrium MC.

Table 3. -- Initial MCs in percent, measured by the cookies.


Mean 57.6 57.2 60.5 75.5 71.5
SD (a) 20.9 22.6 23.7 26.1 29.7
Min. 28.8 32.5 28.8 31.9 34.0
Max. 133.6 184.1 140.3 156.3 172.9

(a) SD = standard deviation.

Table 4. -- Final MC in percent of kiln-dried timbers.

 Control V-ARO Ho-ARO Ho-ARO Ho-ARO

Mean (KD rough) 14.7 15.2 15.8 15.5 15.5
SD (a) 4.3 3.9 7.2 5.5 6.5
Min. 7.1 3.5 7.8 4.5 3.4
Max. 37.3 27.8 51.1 43.1 70.8
Shell (KD rough) 15.9 12.7 19.9 17.5 17.6
SD 5.1 4.0 5.7 6.5 6.5
Min. 9.1 8.0 11.1 9.6 9.0
Max. 40.4 45.4 44.3 60.0 42.9
Core (KD rough) 18.1 13.6 26.3 22.1 21.0
SD 7.5 4.0 9.9 10.2 10.3
Min. 9.1 7.9 10.7 9.9 9.6
Max. 48.0 31.9 60.0 60.0 60.0
 difference 2.2 0.9 6.4 4.7 3.4
SD 3.4 1.9 5.9 5.9 5.4
Min. -1.8 -13.5 -3.1 -3.7 -2.8
Max. 25.6 10.5 33.6 37.0 33.4
Planed KD 17.1 (mean for 5 runs, 800 pieces, in-line meter)

(a) SD = standard deviation.

Table 5. -- Annual ring orientation effect on drying time.

 Control V-ARO Ho-ARO Ho-ARO Ho-ARO

Drying time (days): total time 18.4 25.5 16.3 21.3 22.3
 from 58.4% 15.8 16.9
Drying time change (%) 0 38.6 -11.4 -14.1 -8.2

Table 6. -- Shape distortions measured in the five runs before (green
condition) and after drying.

 After drying
 Before drying (KD rough)
 Twist Bow Diamond Twist Bow Diamond

Control Mean 2.0 2.7 0.3 3.6 3.8 1.4
 SD (b) 1.1 2.5 0.6 2.4 3.0 1.3
 Min. 0.0 0.0 -1.5 0.3 0.4 -0.4
 Max. 4.7 30.4 1.8 17.1 19.8 5.9
V-ARO Mean 2.3 2.5 0.6 3.7 4.3 0.8
 SD 0.9 1.1 0.5 1.7 2.0 0.8
 Min. 0.6 0.0 0.0 0.0 1.1 0.0
 Max. 5.2 6.6 3.0 8.3 12.1 3.6
Ho-ARO Mean 2.0 2.7 0.5 2.8 3.6 0.8
 SD 0.9 1.1 0.4 1.3 1.9 0.9
 Min. 0.0 0.0 0.0 0.2 1.1 0.0
 Max. 5.9 6.4 2.2 7.5 13.9 3.6
H-SOG- Mean 1.6 2.4 0.4 3.4 3.7 1.3
Ho-ARO SD 0.7 1.1 0.3 2.1 2.8 1.3
 Min. 0.1 0.5 0.0 0.0 0.0 0.0
 Max. 3.8 5.4 1.9 13.8 19.2 5.5
L-SOG- Mean 1.8 2.5 0.3 3.6 3.5 1.5
Ho-ARO SD 0.8 1.4 0.4 2.0 2.6 1.3
 Min. 0.3 0.4 0.0 0.4 0.4 0.0
 Max. 4.7 8.7 3.3 11.1 15.2 5.4

 Difference between before
 and after drying
 Twist Bow Diamond

Control Mean 1.6 (81.9) (a) 1.1 (39.9) 1.0 (404.3)
 SD (b) 2.6 3.7 1.4
 Min. -2.0 -26.3 -1.4
 Max. 15.3 15.2 5.5
V-ARO Mean 1.4 (59.9) 1.8 (73.2) 0.2 (43.0)
 SD 1.8 2.2 0.9
 Min. -2.8 -2.7 -2.7
 Max. 7.0 7.8 3.2
Ho-ARO Mean 0.9 (43.5) 1.0 (35.8) 0.3 (62.1)
 SD 1.5 2.0 0.9
 Min. -2.2 -3.3 -1.7
 Max. 5.6 9.2 3.3
H-SOG- Mean 1.8 (109.0) 1.3 (55.6) 0.9 (227.4)
Ho-ARO SD 2.2 2.8 1.3
 Min. -2.1 -3.9 -0.9
 Max. 11.9 18.3 5.2
L-SOG- Mean 1.8 (99.9) 1.0 (38.2) 1.2 (435.4)
Ho-ARO SD 2.2 2.7 1.3
 Min. -2.5 -5.0 -1.6
 Max. 10.0 13.6 5.3

(a) Values in parentheses are percentages.
(b) SD = standard deviation.

Table 7. -- Detailed grading results.

 Clear 1JST & Grade 2 Utility Economy Reject Wet

 No. of pieces 1 56 42 57 4 0
 Percent 0.6 35.0 26.3 35.6 2.5 0.0 0.0
 No. of pieces 1 81 29 49 0 0
 Percent 0.6 50.9 18.2 30.8 0.0 0.0 0.0
 No. of pieces 2 106 21 26 5 0
 Percent 1.2 65.0 12.9 16.0 3.1 0.0 0.0
 No. of pieces 3 53 47 50 7 0
 Percent 1.9 34.0 30.1 32.1 4.5 0.0 0.0
 No. of pieces 10 52 46 46 6 0
 Percent 6.4 33.1 29.3 29.3 3.8 0.0 0.0

 Top 3 grades

 No. of pieces
 Percent 61.9
 No. of pieces
 Percent 69.8
 No. of pieces
 Percent 79.1
 No. of pieces
 Percent 66.0
 No. of pieces
 Percent 68.8

Table 8. -- Dimensions, twist, bow, and diamonding after planing.

 Width Thickness Twist Bow Diamond

 Mean 105.0 105.0 1.7 2.3 0.4
 SD (a) 0.5 0.5 1.0 1.3 0.3
 Min. 103.4 102.6 0.0 0.0 0.0
 Max. 105.9 105.9 6.0 7.2 1.7
 Mean 105.2 105.1 1.7 2.4 0.5
 SD 0.3 0.3 0.8 1.1 0.3
 Min. 104.2 103.5 0.0 0.0 0.0
 Max. 106.6 106.0 4.5 5.9 1.5
 Mean 105.2 105.0 1.8 2.3 0.4
 SD 0.5 0.4 0.9 1.3 0.3
 Min. 103.6 102.7 0.0 0.0 0.0
 Max. 106.4 106.4 5.7 9.5 1.8
 Mean 105.6 105.4 1.7 2.4 0.6
 SD 0.5 0.6 0.8 1.6 0.6
 Min. 103.2 102.4 0.0 0.0 0.0
 Max. 107.1 107.4 4.7 10.0 3.3
 Mean 105.6 105.3 1.7 2.2 0.5
 SD 0.6 0.5 1.0 1.5 0.5
 Min. 103.5 104.2 0.0 0.3 0.0
 Max. 107.8 106.2 5.9 10.7 3.2

(a) SD = standard deviation.

Literature cited

Avramidis, S. 2001. Evaluation of conventional and radio frequency vacuum drying and redrying of pacific coast hemlock hashira and hirakaku timbers. Final report to the stability group, Zairai Lumber Partnership, Vancouver, BC, Canada. 42 pp.

Balodis, V. 1971. Influence of grain angle on twist in seasoned boards. Working group on microscopic characteristic of wood quality. Section 41. IUFRO, Vienna, Austria.

Danborg, F. 1994. Drying properties and visual grading of juvenile wood from the fast grown picea abies and picea sitchensis. Scandinavian J. Forest Res. 9:91-98.

Foresberg, D. and M. Warensjo. 2001. Grain angle variation: A major determinant of twist in sawn picea abies (L.) karst. Scandinavian J. Forest Res. 16:269-277.

Haslett, A.N., I.G. Simpson, and M.O. Kimberley. 1991. Utilization of 25-year-old Pinus radiata. Part 2. Warp of structural timber in drying. New Zealand J. Forest Sci. 21(2/3):228-234.

McCurdy, M.C. and R.B. Keey. 2002. The effects of growth-ring orientation on moisture movement in the high-temperature drying of softwood boards. Holz als Roh-und Werkstoff. 60:363-368.

Pang, S. 2002. Effects of sawing pattern on lumber drying: Model simulation and experimental investigation. Drying Technol. 20(9):1769-1787.

Siau, J.F. 1995. Wood: Influence of moisture on physical properties. Dept. of Wood Science and Forest Products. VPI & SU, Blacksburg, VA. 219 pp.

Bingye Hao

Stavros Avramidis*

The authors are, respectively, Post-Doctoral Research Fellow and Professor, The Univ. of British Columbia, Dept. of Wood Science, 2424 Main Mall, Vancouver, BC, Canada V6T IZ4. This work was financially supported by the Coast Forest and Lumber Association (CFLA) of British Columbia. Our appreciation is extended to Pat Demens of International Forest Products and Bruce St. John of Weyerhaeuser Canada for assisting with specimen acquisition and evaluation, and to Dal Wright of Forintek Canada Corporation for cutting the green specimens. This paper was received for publication in May 2003. Article No. 9675.

*Forest Products Society Member.
COPYRIGHT 2004 Forest Products Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2004 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Hao, Bingye; Avramidis, Stavros
Publication:Forest Products Journal
Geographic Code:1CANA
Date:Nov 1, 2004
Previous Article:Mechanical tree havesters spread fungal inoculum onto freshly felled Canadian and New Zealand pine logs.
Next Article:Diamond wire machining of wood.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters