SciELO - Scientific Electronic Library Online

Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados


Maderas. Ciencia y tecnología

versión On-line ISSN 0718-221X

Maderas, Cienc. tecnol. v.4 n.1 Concepción  2002 

Maderas. Ciencia y tecnología. 4(1):40-49, 2002



S. Pang and A.N. Haslett
1Department of Chemical and Process Engineering and Wood Technology Research Centre,
University of Canterbury, Christchurch, New Zealand.
2Weyerhaeuser Australia Pty Ltd., Mount Gambier, Australia.



Flatsawn and quartersawn Pinus radiata (radiata pine) sapwood boards were dried in a tunnel dryer in three runs using three drying schedules. These schedules represented low temperature (LT), accelerated conventional temperature (ACT) and high temperature (HT) drying. In each run, the boards were assembled in the dryer in two vertically matched layers with the flatsawn and the quartersawn boards being alternatively placed side by side. The samples were weighed at pre-set drying times. After drying, all of the samples were cut to two halves, one half for oven drying to determine the moisture content during drying and the remaining half for stress assessments. For LT and ACT drying, the quartersawn boards dried slower than the flatsawn boards, but for HT drying the difference between the quartersawn and the flatsawn boards was not statistically significant. The transverse residual drying stress in the quartersawn boards was, in general, lower than the flatsawn boards and the difference was most apparent for the LT drying.

Keywords: drying rate, drying schedule, drying stress, flatsawn, kiln drying, lumber, Pinus radiata, quartersawn, sawing pattern


In the kiln drying of softwood lumber, there are two major concerns regarding the dry lumber quality. The first one is the residual drying stresses and board distortion, and the second one is the moisture content variation, both across the dried stack and within a single piece. These drying concerns are strongly related to the variation of wood properties, and become more pronounced for fast growing plantation species.

Due to differences in wood permeability and shrinkage coefficients between the tangential and the radial directions, the sawing pattern of lumber affects drying rate and drying stresses. Within commercial kiln drying, stack of the lumber always shows varying sawing patterns although, due to the grade sawing around the log, the majority of the Pinus radiata boards sawn in New Zealand are predominantly tangentially sawn. During drying, the liquid moisture and the moisture vapour within the wood material can either flow along the radial ray cells or through bordered pits which are on the radial-longitudinal face of tracheids. Whereas the bordered pits aspirate with drying, the ray cells are cleared. This increases the gas permeability and the moisture flow in radial direction (Booker and Evans, 1994).

Due to non-uniform shrinkage of wood during drying, drying stress develops and its gradient in the board thickness is the highest because the drying mainly occurs from the faces of the board. Previous studies by Pang (2001) have shown that the tangential shrinkage coefficient is about 30% higher than the radial shrinkage coefficient for Pinus radiata, thus the residual drying stresses in the board width direction can also vary with the board sawing pattern. The objective of the current study was to quantify the effects of the sawing pattern on the drying rate and the residual drying stresses.


For preparation of the drying samples, two straight, pruned butt logs were selected from a forest in the Central North Island, New Zealand. The logs were 5 m long and had a diameter of 55-60 cm, with the pith being centrally located at each end. The logs were then sawn to 100´ 100 mm squares using a cutting pattern as shown in Figure 1. Half of the sapwood squares were ripped into two, 100´ 40 mm tangentially sawn (flatsawn) boards and the remaining half of the sapwood squares were sawn into two, 100´ 40 mm quartersawn boards. Ideally the flatsawn board had a zero growth ring angle to the board face and the quartersawn board had a growth ring angle of 90°. However, it is extremely difficult to cut the precisely flatsawn or quartersawn material using industrial band saws, and consequently, the experimental flatsawn boards had growth ring angles varying from 0 to 40° and the quartersawn boards had growth ring angles from 50 to 90°. After sawing, the 100´ 40´ 5000 mm boards were crosscut into 580 mm long drying samples. During cutting, 25 mm cross sections were cut from each end of the samples to determine the green moisture content and the wood basic density.

Figure 1. Generalised sawing pattern to obtain 100’ 100 mm squares.

The 100’ 40 580 mm sapwood samples were dried in a tunnel dryer using three drying schedules (three runs). The dryer had a drying chamber of 600 mm long, 200 mm high and 3.5 m wide, and operates from 45°C up to 190°C. Airflow of up to 9 m/s can be achieved and the flow direction can be revered as required.

The three drying schedules used include low temperature (LT) drying, accelerated conventional temperature (ACT) drying and high temperature (HT) kiln drying. The drying schedule and the air conditions for each run are given in Table 1. In each run, 40 selected samples, 20 flatsawn and 20 quartersawn, were assembled in the drying tunnel in two layers, 20 boards in each layer and vertically matched with the same growth ring angle. In each layer, the flatsawn and the quartersawn boards were alternatively placed side by side across the stack. Supplementary sapwood boards were also placed in blank layers above and below the two test layers. During the course of drying, boards from the middle two test layers were weighed at pre-set drying times (Table 1). After drying, all of the test samples were re-weighed and each board was then cut into two 290 mm lengths. One length was oven-dried for determining the moisture content during drying and the remaining length was used for assessing the residual drying stresses. The growth ring angle for each sample board was measured as an indicator of the sawing pattern. Because the growth ring angle varied over the cross section of the board, the growth ring angle was rounded to the nearest 10°.

For assessment of the residual drying stresses, two 25 mm cross sections were cut from each test sample and each section was rippled at mid-thickness into two halves as shown in Figure 2. Then, the cup of each half piece (D ) was measured by using a dial gauge fixed on a straight edge (accuracy of 0.01 mm).

Figure 2. Mid-thickness ripping of a cross section for cup measurement (residual drying stress).

Table 1. Drying schedules and pre-set drying times before sample weighing.



Temperature (°C)


Drying time for samples weighing




Reversal (h)







8, 16, 22, 30.5, 38, 45, 54, 61, 67, 77







4, 8, 12, 17, 24.5, 28.8, 33.3, 39.7, 48.6, 52.4







2.5, 5, 7.5, 10, 12.5, 15.5

1: dry-bulb temperature; 2: wet-bulb temperature.


Drying rate

In addition to sawing pattern, variation in green moisture content, wood density and air conditions can also affect the drying rate and the final moisture content distribution as studied by Pang (2002). However, if the green moisture content and the wood density have a scattered distribution against the growth ring angle, their influence can be ignored when investigating the effects of the sawing pattern. When the green moisture content of all the test samples are plotted against the growth ring angle (Figure 3), it can be seen that the green moisture contents were scattered from 150% to 210%, but most of the values fell in a range between 160% and 200%. Regression analyses showed that there was no significant relationship between the green moisture content and the growth ring angle. Further examination neither showed any significant relationship between wood density and the growth ring angle.

Early studies by Pang (1994) have shown that during drying the air temperature decreases and the air humidity increases along the airflow direction. The air property changes are most apparent in initial half of the drying, and the changes become less important towards the end of drying. In order to minimise the influence of the air condition changes, the airflow direction is reversed periodically. In addition, in the current studies the quartersawn boards and the flatsawn boards were arranged alternatively side by side in each of the two test layer.

Figure 4 shows the moisture content of individual boards at different times during ACT drying. Before drying, the moisture content of the boards was randomly distributed (R2=0.05), ranging from 148% to 195% with an average value of 170%. This random distribution was maintained during the first 12 hours of drying in which the average moisture content decreased to 90%. During this period of time, the wettest boards remained the wettest and the driest boards remained the driest. However, after 17 hours of drying during which the average moisture content fell to 69%, a trend was observed that the moisture content increased with the growth ring angle (R2=0.43). This indicates that the quartersawn boards started to dry more slower than the flatsawn boards. This trend became clearer with further drying, R2 being 0.63 after 33.3 hours and 0.73 after 48.6 hours of drying. The average moisture content at the two corresponding drying times were 25% and 9%, respectively.

In LT drying, the moisture content distribution during drying was similar to that for the ACT drying. In LT drying, the trend for the quartersawn boards drying more slower than the quantersawn boards appeared after 38 hours of drying. At this time the average moisture content was 50%. As observed in ACT drying, the difference between the quartersawn boards and the flatsawn boards became more pronounced with further drying. At the end of drying, after 77 hours, when the average moisture content was 13%, the moisture content was higher for the boards with higher growth ring angles (R2 value being 0.56 as shown in Figure 5).

Interestingly, during HT drying there was no trend was observed regarding the drying differences between the quartersawn boards and the flatsawn boards. Through the drying course, the relationship between the moisture content and the growth ring angle was insignificant with R2 values being less than 0.1. After 15.5 hours of drying, the average moisture content was 8.6% and the moisture content distribution as a function of growth ring angle is shown in Figure 6. In addition, the moisture content after HT drying was more variable than after the LT drying and the ACT drying, indicating the air condition variation across the stack may have played a more important role in the HT drying than the growth ring angle.

The effect of growth ring angle on the drying rate is attributed to transverse permeability differences. In green wood, the liquid water permeability in tangential direction is higher than that in the radial direction (Booker 1991). However, at high moisture content when the surface is saturated, the liquid flows from the cell lumens towards the board surface due to a capillary force and the drying occurs near the exposed surfaces. During this period, the drying is controlled by the external conditions and a thin drying shell develops around the board surfaces (Pang et al. 1994). Therefore, when the moisture content is high, liquid permeability has only a limited influence on the drying and consequently the sawing pattern does not have significant impact on the drying rate.

Figure 3. Distribution of green moisture content (shows no relationship with the growth ring angle).

Figure 4. Moisture content of individual boards as a function of growth ring angle during ACT drying (Run 2).

Figure 5. Moisture content of individual boards as a function of growth ring angle after 77 hours of LT drying (Run 2).

Figure 6. Moisture content of individual boards as a function of growth ring angle after 15.5 hours of HT drying (Run 3).

When the surface moisture content falls below the fibre saturation point, the water evaporation occurs at positions further from the board surfaces, thus the drying becomes dominated by the vapour diffusion within the wood. In the subsequent drying, gas permeability has a significant influence on the drying rate. Within the dry zones of the board, as the radial ray cells are partly cleared of resin and extractives, and the bordered pits have aspirated during drying, the gas permeability in the radial direction is higher than that in the tangential direction (Booker 1991; Booker and Evans 1994). In LT and ACT drying, the gas permeability difference between the radial direction and the tangential direction is the dominant factor affecting the drying rate once the board surface is below the fibre saturation point. Because the drying mainly occurs along the thickness direction of a board, the flatsawn board with its thickness parallel to radial direction dries faster than the quartersawn board (its thickness is in the tangential direction).

However, in HT drying the wood temperature can reach or exceed the water-boiling point, meaning that the vapour pressure can increase abruptly with a slight increase in the wood temperature. In this case the wood temperature plays a more important role than the wood gas permeability. This can explain the scattered moisture content distribution observed in the HT drying.

Residual drying stress

After drying, the board surface is normally in compression and the core in tension. Mid-thickness ripping removes the restraint between the two halves (Figure 2), and each half bends inward due to a bending momentum generated by the thickness stress gradient. The higher the transverse stress is, the more each ripped half will bend (cup). Therefore, the average cup of the two halves is taken as an indicator for the transverse residual stresses after drying. The measured cup values from the dried boards for the three runs are shown in Figures 7 (LT drying), Figure 8 (ACT drying) and Figure 9 (HT drying).

Figure 7. Average cup values of two halves for individual boards as a function of growth ring angle after LT drying (Run 1).

Figure 8. Average cup values of two halves for individual boards as a function of growth ring angle after ACT drying (Run 2).

Figure 9. Average cup values of two halves for individual boards as a function of growth ring angle after HT drying (Run 3).

In general, the cup values increased with drying temperature, the overall average cup values for LT, ACT and HT drying were, respectively, 0.53 mm (± 0.31 mm), 0.69 mm (± 0.17 mm) and 0.93 mm (± 0.16 mm). The average cup for the quartersawn boards was generally lower than that of the flatsawn boards. This trend was most apparent for LT drying as shown in Figure 7 with a R2 of 0.76. For LT drying, the quartersawn boards had an average cup of 0.25 mm (± 0.14 mm) whereas the flatsawn boards had an average cup of 0.81 mm (± 0.10 mm). For ACT and HT drying, the relationship between the cup and the growth ring angle was not as strong as in LT drying. For ACT drying, the average cup for the quartersawn boards was 0.65 mm (± 0.18 mm), only 0.08 mm less than that of the flatsawn boards (0.73± 0.16 mm). For HT drying the average cup for the quartersawn boards was 0.86 mm (± 0.17 mm) and that of the flatsawn boards was 1.00 mm (± 0.11 mm).

During drying, internal stresses are induced by differential shrinkage of wood which in turn is caused by moisture content gradient and/or the wood shrinkage coefficient (Pang 2000; Pang 2001). In the early stage of drying when the surface moisture content falls below the fibre saturation point, the shell of the board attempts to shrink but the core is still wet and does not shrink. For the forces within the board to be balanced, the surface shell is in tension and the core in compression. Because the shell is much thinner than the core, the tensile stress in the shell is much higher than the compressive stress in the core. With the drying progressing, permanent set is generated at the board surface due to wood rheological behaviour and the material yield (if the stress is high enough). The rheological behaviour (mechano-sorptive strain and creep) is a well-known property of wood which induces prolonged deformation of stressed wood with changing moisture content. Material yield occurs when the stress is higher than the elastic limit of the material. The permanent set at the board surface prevents deformation recovery of the wood when the core is dried to below the fibre saturation point and tries to shrink. This will reverse the stress pattern in the late stage of drying and result in a compressive stress at the board surface and a tensile stress in the core.

Because the radial shrinkage coefficient is about one third lower than the tangential shrinkage coefficient (Pang, 2001), the quartersawn board shrinks less in its width direction than that of the flatsawn boards. Consequently, after drying the quartersawn board has less shrinkage differences and less residual stresses than the flatsawn board. This is more apparent for LT drying because the mecahno-sorptive strain and the creep are less at low wood temperatures, thus the shrinkage coefficient differences play a more important role in the stress development than in the ACT drying and the HT drying. In addition, for LT drying the stress level in the early stage of drying may not reach the yield point of the material which reduces stress reversal.

For drying with higher temperatures (ACT and HT), the moisture content gradient is greater and the drying stresses are higher than with low temperature drying. In addition, with higher temperature drying the wood temperature is higher thus the mechano-sorptive strain and the creep are greater than in LT drying. All of these factors contribute to a higher level of stresses in the late stage of ACT and HT drying. These factors have reduced the influence of the wood anisotropic shrinkage differences between the quartersawn boards and the flatsawn boards as observed in Run 2 and Run 3.


In the kiln drying of lumber using low temperature (LT) or accelerated conventional temperature (ACT) schedules, the mixed charges of quartersawn boards and flatsawn boards show higher drying variations than a charge with entirely quartersawn boards, or a charge with entirely flatsawn boards. In the same stack, in the late stage of drying the quartersawn boards dry slower than the flatsawn board and thus the quartersawn boards tend to be under-dried and the flatsawn boards tend to be over-dried. However, this difference is not significant in high temperature drying. The transverse residual drying stresses increase with drying temperatures and the sawing pattern also has some influence on the stresses, the latter influence being more significant for low temperature drying with which the residual drying stresses in quatersawn boards are less than those in flatsawn boards.

The current study shows that the final moisture-content variation of kiln dried pine boards could be reduced by segregation of the quartersawn boards and the flatsawn boards. Using low temperatures (70/50°C), the drying of sapwood quartersawn boards of radiata pine needs about 10 hours longer than the drying of flatsawn boards. The difference in the drying time for ACT (90/60°C) drying is about six hours, however, the drying time difference is statistically negligible for HT (140/90°C) drying. For high quality product, it is suggested to cut the lumber as quartersawn and to dry the lumber using low temperature schedule.


The experimental work was performed at Forest Research, New Zealand. The authors acknowledge the financial support by the New Zealand Foundation of Research, Science and Technology.


Booker, R.E. 1991. Changes in transverse wood permeability during the drying of Dacrydium cupressinum and Pinus radiata. New Zealand J. Forestry Sci. 20(2): 231-244.

Booker, R.E. and Evans, J.M. 1994. The effect of drying schedule on the radial permeability of Pinus radiata D. Don. Holz als Roh- und Werkstoff 52: 150-156.

Pang, S. 1994. High-Temperature Drying of Pinus radiata Boards in a Batch Kiln. Ph.D. Thesis. University of Canterbury, Christchurch, New Zealand.

Pang, S. 2000. Modelling of stress development during drying and relief during steaming in Pinus radiata lumber. Drying Technol. 18(8): 1677-1696.

Pang, S. 2001. Anisotropic shrinkage, equilibrium moisture content and fibre saturation point of earlywood and latewood of radiata pine. Proceedings 7th IUFRO International Wood Drying Conference, Tsukuba, Japan: 184-191.

Pang, S. 2002. Applications of mathematical models to investigate effects of wood variability and rheological properties on lumber drying. Chemical Engineering Journal, 86(1-2): 103-110.

Pang, S.; Langrish, T.A.G. and Keey, R.B. 1994. Moisture movement in softwood timber at elevated temperatures. Drying Technol. 12(8): 1897-1914.

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons