Simulation of the Radial Sawing Technique for Pedunculata Oak (Quercus robur L.) Logs

Abstract

Using the RadSawSim simulator for radial sawing, a simulation of the radial sawing technique was used to saw Pedunculata oak (Quercus robur L.) logs. Simulation was implemented with a view to producing as many radially sawn boards as possible and took into account the influences of increasing volume yield, lumber value yield, and log-value yield. The methods that were analyzed were live sawing and radial sawing of third sections, fourth sections, fifth sections, and sixth sections of the log. Live sawing achieved the best results of volume yield during simulation, which was followed by radial sawing into the third, fourth, fifth, and sixth sections. The difference in volume yield with live sawing compared to the radial-sawing method is very large for logs up to a diameter of 45 cm. It becomes smaller when the log diameter is greater than 45 cm. A comparison of the radial method shows that the share of radially sawn boards and lumber value yield increased when the number of log sections during sawing simulation increased. If log-value yield is assumed to be the main criterion, and given the conditions used in this simulation, there is no justified reason to saw logs using the radial technique when the diameter is less than 45 cm. The live sawing technique is more efficient for these diameters of logs, and, therefore, the radial sawing technique is more efficient for logs with a diameter greater than 45 cm.

1. Introduction

Production of wood products with radial texture in Croatia can be traced back to when oak forests in the Slavonia region (Eastern Croatia) were first exploited for the production of barrel staves at the beginning of the 19th century. At that time, staves were made by hewing logs. Specific sawing techniques used to produce boards with a radial texture began with an increase in demand for oak boards for the British market at the end of the 19th century [1]. Quarter sawing and Slavonian sawing were the primary methods used by French entrepreneurs in Slavonian sawmills. In the Slavonian method of sawing, the log is first sawed in half with simultaneous production of one or more center planks with a radial texture. Log halves are further sawed, removing only side parts in order to make a compact-edged half-log or resawn-edged half-log [2,3]. From experience, written records, and photographs, it is evident that these techniques were used on Pedunculate oak (Quercus robur L.) and Sessile oak (Quercus petraea Matt.), Common beech (Fagus sylvatica L.), and European silver fir (Abies alba Mill.) logs with large diameters in Croatian sawmills.

Even though there is a large demand for oak wood products with radial texture, to date, the radial sawing technique on oak logs has not remained dominant, and other sawing techniques, like live sawing, prevail. The main reason for this is the complexity and high cost of radial sawing compared to other simpler methods of sawing [4].

Even though research on European silver fir was conducted [5], the comparison of quarter sawing and cant sawing shows positive financial results when better-quality logs were sawn, primarily using the butt log. With respect to the average board width and volume yield, quarter sawing produces somewhat poorer results than cant sawing. Production of radially sawn boards with these types of logs decreases the wood defects found in the butt swell. During the drying process, radially sawn boards also do not deform as much and are more valuable for the production of final wood products [6,7].

The selection of appropriate sawing patterns plays a crucial role in determining both the quality and yield of beech (Fagus sylvatica L.) lumber. Vilkovský et al. demonstrated that, while cant sawing yielded the highest quantitative output (up to 84%), quarter sawing produced superior qualitative results due to the higher proportion of radial boards, which offer enhanced dimensional stability and reduced warping during drying [8]. Similarly, Klement et al. [9] found that the orientation of sawing significantly affects the performance of beech blanks, with boards sawn parallel to the timber’s axis showing higher yield values (65.14%–72.70%) and less longitudinal warping than those sawn parallel to the edge (60.38%–68.40%). These findings underscore the importance of aligning the sawing strategy with desired product quality [9]. Additionally, research indicates that, while log dimensions have limited influence on yield, the chosen sawing method, such as radial or live sawing of half logs, significantly affects both yield and board quality [10]. Therefore, optimizing sawing patterns and orientation is essential for maximizing the value and performance of beech wood products.

Keeping in mind the preferences regarding sawing products with radial textures, Sandberg researched specific star-sawing methods of radial-sawing sixth sections that are used to produce wood elements for glued laminated timber and solid wood panels. Simulation results showed a very acceptable volume yield in the form of sawn boards of 65%–70% [11]. In further research, Sandberg determined a series of advantages in wood products with radial texture on Scots pine (Pinus sylvestris L.) and Norway spruce (Picea abies Karst.) and developed a new production method (PrimWood Method) for high-quality glued laminated timbers [12].

One of the practical examples of producing radially sawn boards is found in Australia. The technology machines round wood of smaller diameters with the goal of achieving a high-volume yield. The production line is managed by a computer. Machine processing begins with removing the bark with the debarker, and then the log is scanned using a laser. The computer then determines the optimal sawing position of the log using the circular log saw that saws the log into third, fourth, fifth, sixth, or eighth sections. Each section is then individually sawn into sawn boards with the multiple circular saw and chipper profiler [13].

Begunkova [14] proposed a predictive method for estimating the yield of radially sawn timber from round logs using a texture-modeling software system. The approach is based on numerically evaluating the linearity and orientation of annual rings to classify board surfaces as radial, semi-radial, or tangential. The study demonstrated that key influencing factors on radial yield include log diameter, the shape of the log’s generatrix (e.g., straight line, parabola, or Neil’s parabola), and the direction of sawing (parallel to the axis or to the generatrix). Results showed that sawing parallel to the generatrix with additional longitudinal board cutting significantly increases the radial yield, especially when using logs with a straight generatrix. The developed methodology allows for simulation-based prediction and optimization of sawing patterns to maximize the output of high-quality radial boards, which is especially relevant for the production of dimensionally stable and visually uniform wood components [14].

The goal of this research was to analyze the methods of sawing Pedunculate oak logs with the radial technique using the computerized simulator for radial sawing—RadSawSim. An analysis was conducted with the view of producing as many radially sawn, high-quality boards as possible. The following sawing methods were analyzed:

• live sawing;
• radial sawing third sections;
• radial sawing fourth sections;
• radial sawing fifth sections;
• radial sawing sixth sections.

The criteria for choosing the most effective sawing method were the share of radial, half-radial, and flat-sawn boards and the effects of differing log diameters on the volume, lumber value yield, and log-value yield when producing sawn boards of various thicknesses. Log volume yield, lumber value yield, and log-value yield were calculated according to Ištvanić et al. [15]. The oak was chosen as the object of this research due to the well-known value of radial and half-radial oak sawn boards on the European wood market. Transferring this sawing simulation to other wood species is not a problem, and the characteristic parameters for the wood species in question would be inserted in the mathematical simulation model.

2. Materials and Methods

More information on sawing methods that produce sawn boards with a radial texture presents a significant problem in Croatian sawmills since they are rarely used and since the current technology is adjusted to the live- or cant-sawing methods. A series of constraints further limit the organization of the actual test sawing that monitors the log yield. Experimental test sawing is expensive, time-consuming, and requires good organization and professional staff who prepare the raw material and then process and interpret the data.

With today’s level of computer technology and the use of specialized software based on mathematical models, it is possible to substitute experimental with simulation sawing. This research implemented the RadSawSim (v1) simulator for radial sawing that was developed by the Faculty of Forestry at the University of Zagreb [15]. The installation method of the simulation and its limitations are described below.

2.1. Description of the Simulated Method of Sawing

Live sawing is also known as slab sawing or through-and-through sawing. Live sawing is a traditional method of cutting wood in which the log is sawn into sawn boards by passing the saw straight through the log. The result is boards with a mixture of tangentially, semi-tangentially, or semi-radially and radially sawn texture patterns (Figure 1a). Live sawing is a well-known method that is very often used in European and Croatian sawmills, and, in this research, it is used to compare results with other sawing simulation methods. Radially sawing third sections is not commonly used in Croatian sawmills. However, sawing logs into thirds (flitching) exists and is used to produce cut veneer. This is why it is used in this research as a new method of log sawing and sawing into sawn boards. In order to obtain three-thirds of the log, the log must first be sawn in half along its length, and then one-sixth needs to be sawn from each log half. This approach has constraints since adjustments to the customary technological base and machines used in Croatian sawmills are needed. Radial sawing fourth sections (quarter sawing) is known in theory but is rarely used in Croatian sawmills [3]. Quarter sawing first saws the log in half along its length, then each half is sawn into fourths.

As in radially sawing third sections, radially sawing fifth sections is not commonly used in Croatian sawmills. In this case, the logs are sawn in half along the length. Each half is then sawed into two-fifth sections and one-tenth sections. In this manner, we receive a four-fifth section and two-tenth sections which together make up log fifths. As in sawing third sections, the technology and machinery normally used in Croatian sawmills would need to be adjusted. Like radially sawing third and fifth sections, the radial sawing of sixth sections is also not commonly used in Croatian sawmills. In this case, we first assume that the log will be sawn in half along its length, and then each half will be sawn into three-sixths sections.

The third, fourth, fifth, and sixth sections are then sawn into an optimal even and odd number of sawn boards, as shown in Figure 1b–e.

2.2. Simulation Parameters

For simulation, the mid diameter varied from 25 to 100 cm with an increase of 5 cm for each test. This means that 16 log diameters were monitored for each sawing method and for each sawn board thickness, which totaled 320 simulated logs. Even though the mathematical simulation model also foresees the possibility of varying the log’s length and log taper, in this research, they were used as constants: the length of the log was 4 m, and the log taper was 1.5 cm/m′. The simulation hypothesized the production of edged sawn boards of nominal thicknesses of 25, 32, 38, and 50 mm. In the simulation, the entire length of the log was always sawn into sawn boards of varying thicknesses. The model optimizes sawing to incorporate the maximum number of even and odd-numbered sawn boards, given the specified sawn-board parameter. The minimal nominal width of the sawn boards was 8 cm, and nominal dimensions assumed 12% moisture content. Only sawn boards from the center part of the log that had the minimal nominal width, and the log length was monitored. Sawn boards from the outer portion of the log that did not have the minimal nominal width and log length were not hypothesized in the simulation.

Given that the method of sawing requires technological adjustments for conditions in Croatian sawmills, this simulation uses the log band saw. The mathematical model also foresees the possibility of varying the sawkerf width that can accommodate most modern sawing machinery. However, in this research, the sawkerf is a constant of 3 mm. The lumber value yield calculation for sawn boards’ quality is presumed to be based on the market value of radially sawn boards, with half-radial and tangent texture added to the quality index. This simulation of radial sawing used the value 1 for radially sawn boards, 0.9 for half-radially sawn boards, and 0.8 for tangentially sawn boards [15].

3. Results and Discussion

3.1. Share of Radially Sawn Boards

As expected and shown in Figure 2, the radial method, compared to live sawing, produces more sawn boards with radial texture. A comparison of simulation radial methods shows that the number of sawn boards increases as the number of sections of the log increases. Taking into consideration all simulation sawn-board thicknesses, the share of radially sawn boards with live sawing was approximately 45% of the total sawn-board volume. The share ranged from 40% to 100% for sawing third sections. The share for sawing fourth and fifth sections was 60%–100% and for sawing sixth sections it was 80%–100%.

3.2. Log Volume Yield

During simulation, the results of volume yield for logs with a diameter of 25 to 100 cm and production of sawn boards with a thickness of 25, 32, 38, and 50 mm for simulated methods of sawing are shown in Figure 3.

It is evident that the best yield was produced by the simulation of live sawing, followed by radial sawing third sections, and then by radial sawing fourth sections, fifth sections, and lastly, sixth sections. The difference in live sawing compared to the radial method is largest for log diameters up to 45 cm, after which the difference reduces. During simulation of live sawing, the volume yield increases from 0.60 to nearly 0.75 m³ sawn boards per m³ log. Volume yield for simulated radial sawing shows significantly lower results, particularly when logs up to a 50 cm diameter were examined. In this simulation case, results show the questionable rationale for sawing logs of this diameter type using the radial method, primarily from the viewpoint of the volume yield. The volume yield amounts to 0.5 to 0.7 m³ sawn boards per m³ log for logs with a diameter greater than 50 cm, depending on the method of radial sawing and simulated sawn-board thickness.

3.3. Lumber Value Yield

Lumber value yield results during the simulated sawing of logs with a diameter ranging from 25 to 100 cm and production of sawn boards with a thickness of 25, 32, 38, and 50 mm for the sawing methods used in this research are shown in Figure 4. The figure clearly shows the trend that lumber value yield follows the ratio of sawn boards with radial texture. The smallest lumber value yield was achieved with simulated live sawing, followed by radial sawing into third, fourth, and fifth sections. The greatest lumber value yield was achieved with radial-sawing sixth sections.

3.4. Log-Value Yield

The log-value yield results obtained from the simulated sawing of logs with diameters ranging from 25 to 100 cm, and the production of sawn boards with thicknesses of 25, 32, 38, and 50 mm, are presented in Figure 5. As shown in Figure 5, as well as in

Table 1. Best log-value yield results given simulation of mid-diameter logs, nominal sawn board thickness, and method of sawing.

Mid Diameter (cm)	Board Thickness (mm)
	25	32	38	50
25	n2 0.521	n2 0.560	n2 0.524	n2 0.456
30	n2 0.515	n2 0.541	n2 0.568	n2 0.522
35	n2 0.567	n2 0.548	n2 0.550	n2 0.522
40	n3 0.568	n2 0.560	n2 0.579	n2 0.570
45	n4 0.578	n2 0.580	n4 0.572	n3 0.563
50	n4 0.592	n3 0.582	n4 0.586	n2 0.560
55	n4 0.610	n3 0.609	n3 0.605	n4 0.589
60	n6 0.609	n5 0.622	n4 0.621	n3 0.586
65	n6 0.619	n4 0.638	n3 0.614	n3 0.629
70	n4 0.648	n4 and n6 0.629	n4 0.638	n4 0.642
75	n5 0.642	n4 and n5 0.630	n4 0.640	n4 0.639
80	n4 0.645	n4, n5 and n6 0.635	n4 and n6 0.642	n3 0.638
85	n6 0.655	n4 0.675	n4 0.661	n3 0.640
90	n4 0.658	n5 0.670	n5 0.652	n4 0.650
95	n5 0.655	n4 0.676	n3 0.660	n4 0.660
100	n6 0.665	n5 and n6 0.678	n4 0.680	n4 0.658

Note. Method of sawing: n2 = live sawing; n3 = radial sawing thirds sections; n4 = radial sawing fourths sections; n5 = radial sawing fifths sections; and n6 = radial sawing sixths sections of log.

, live sawing yielded the most favorable results for logs with diameters up to 45 cm. The simulation further indicates that logs with diameters of 25 and 45 cm are most suitable for producing 32 mm-thick sawn boards. Logs with diameters of 30 and 40 cm are optimal for producing 38 mm-thick boards, while logs with a diameter of 35 cm are best suited for the production of 25 mm-thick boards.

Log diameters greater than 45 cm showed better results during the simulation of the radial method. Simulation showed that radial sawing fourth sections from logs with a diameter of 50, 55, 70, and 80 cm best produces 25 mm-thick sawn boards, while the live sawing method produced the worst average result. Logs with a diameter of 65, 85, and 95 cm most effectively produce 32 mm sawn boards, and logs with a diameter of 75 and 100 cm best produce 38 mm sawn boards. Radial sawing of fifth sections from logs with a diameter of 60 and 90 cm best produces 32 mm sawn boards.

Table 2. Bayesian Estimates of Value Yield for Different Sawing Methods.

Method of Sawing	Posterior Mean (Value Yield)	95% Credible Interval Lower	95% Credible Interval Upper
Live sawing (n2)	0.544	0.529	0.558
Radial sawing—thirds (n3)	0.609	0.590	0.627
Radial sawing—fourths (n4)	0.634	0.622	0.647
Radial sawing—fifths (n5)	0.645	0.621	0.670
Radial sawing—sixths (n6)	0.645	0.618	0.672

presents the Bayesian estimates of coefficients for the effect of sawing method on value yield. The posterior means and corresponding 95% credible intervals clearly indicate differences among the applied sawing techniques. The lowest value yield was obtained for live sawing (mean = 0.544; 95% CI: 0.529–0.558). In contrast, all radial-sawing methods showed substantially higher yields: thirds sections (mean = 0.609; 95% CI: 0.590–0.627), fourths sections (mean = 0.634; 95% CI: 0.622–0.647), fifths sections (mean = 0.645; 95% CI: 0.621–0.670), and sixths sections (mean = 0.645; 95% CI: 0.618–0.672). The results demonstrate that radial sawing consistently produces higher recovery compared to live sawing. Differences among the various radial-sawing subdivisions are relatively small, with overlapping credible intervals, suggesting that the choice of subdivision level (thirds, fourths, fifths, and sixths) has less influence on yield than the fundamental distinction between radial and live sawing.

4. Discussion

The simulation results closely align with previous studies in industrial sawmilling, demonstrating reliable predictions of volume yield. For logs sawn into 50 mm boards, live-sawing volume yields were comparable to earlier reports [8]. Average yields for boards of 25, 38, and 50 mm-thickness were higher than previously documented for oak, while smaller-diameter logs (30–35 mm) produced yields consistent with prior findings [16]. As expected, a gradual decrease in volume yield was observed with decreasing log diameter, corroborating established trends for live sawing and thinner boards [17]. Moreover, compared to other studies performed on hardwoods, live sawing consistently produced a greater number of thinner boards compared to radial and other sawing methods [18], while maintaining the lowest proportion of radial boards, as also reported in earlier research [19]. These observations confirm the simulation’s accuracy and highlight the characteristic performance of sawing methods across different log sizes and board thicknesses, offering valuable insights for optimizing sawmilling strategies. The higher average volume yields observed for boards of 25, 38, and 50 mm-thickness may be attributed to the optimized log parameters and sawing sequences used in the simulation, which likely reduced wood loss compared to traditional industrial practices. Additionally, differences in log quality, taper, or assumed sawing parameters between studies could contribute to the observed discrepancies, highlighting the importance of considering log-specific characteristics when evaluating sawing efficiency.

5. Conclusions

Simulated radial sawing shows that volume yield decreases when the number of sections increases for the examined log diameters; however, the share of radially sawn boards and the lumber value yield increases. The difference between log-volume yields during live sawing compared to the radial-sawing method is very large with logs of smaller diameters; however, the difference diminishes with the increase in log diameter. If we assume log-value yield is the main criterion for determining the sawing method, there is no justification for sawing logs with the radial technique when the diameter is less than 45 cm. The live-sawing method would be better for these types of logs. For logs with a diameter greater than 45 cm, the radial sawing technique is more effective. When examining the radial technique, the best results were produced when sawing fourth and fifth sections. Therefore, it is best to produce sawn boards with a thickness of 25, 32, and 38 mm, depending on the log diameter. Further steps in this research should confirm the results of this simulation with experimental sawing at a Croatian sawmill. In this way, the simulation and experimental results could be compared, which would provide a proper conclusion. The results obtained from the sawing simulations conducted in this study, when considered alongside existing research on the implementation of 3D computed tomography (CT) scanners in sawmills [20,21], suggest a significant potential for improving the efficiency and optimization of log breakdown. The integration of advanced simulation models with real-time internal log structure analysis provided by CT-scanning technology could enhance decision-making in sawing strategies, ultimately leading to increased yield, reduced waste, and more precise value recovery from logs. This potential should be further explored in subsequent research to validate its practical implementation and optimize its effectiveness across different log classes and sawing methods.