Charring Properties of Korean Larch Structural Glue-Laminated Timber Beams Based on Cross-Sectional Area Ratios

Abstract

Carbon emissions accelerate global warming and climate change, prompting the global development of strategies for carbon reduction. Wood, with its excellent carbon storage capacity, is a sustainable and environmentally friendly material. One cubic meter of timber can absorb 1 t of carbon dioxide and store 250 kg of carbon. This study aimed to conduct fire resistance tests on structural glue-laminated timber beams made from Korean larch (Larix kaempferi) and analyze their char properties. The specimens were fabricated with different cross-sectional shapes and areas and underwent load-bearing fire resistance tests. The results were analyzed in terms of char depth, char rate, and changes in char thickness based on the aspect ratio of the beams. In the smaller specimens, the char properties were influenced more by the width than by the length of the beam. Additionally, at a constant cross-sectional area, charring was deeper when the width was shorter than the height. The specimens did not exhibit significant differences in displacement behavior, with all specimens displaying displacements below the maximum permissible value, indicating suitable fire resistance. The findings of this study provide a foundation for research and development of fire resistance design standards for wooden structures utilizing Korean timber.

1. Introduction

Owing to the rise of various environmental concerns, such as global warming and climate change, the use of timber in construction and the requirement for wooden structures have increased significantly. The ability of timber to store carbon and its lower carbon dioxide emissions compared to concrete or steel have positioned it as a sustainable building material with excellent carbon reduction potential. However, it is crucial to acknowledge that even within the same species, timber exhibits variations in its physical characteristics based on the local climate and surroundings [1]. Currently, approximately two-thirds of the timber in Korea consists of older trees that have diminished carbon absorption capacity [2]. To reduce carbon emissions and promote the circulation of timber resources, there is an urgent need to utilize Korean timber, which has strong carbon storage potential. In the construction sector, wood should be employed in the structural elements of buildings to facilitate this resource circulation. One cubic meter of timber can absorb 1 t of carbon dioxide and store 250 kg of carbon. Additionally, timber is highly competitive as a structural material, offering superior tensile strength, compressive strength, and flexural strength per unit weight compared to steel or concrete. Furthermore, timber provides excellent thermal insulation, with insulating performance that is 10 times better than concrete and 500 times better than steel [3,4]. As a result, an increasing number of diverse structures are being constructed with timber worldwide, leveraging these advantageous characteristics.

Wood is a natural material that can exhibit physical variations even within the same species, influenced by the region or environment during growth. To control these physical properties and ensure quality as a building material, raw wood is processed for structural use, resulting in products known as “engineered wood”. Two major forms of engineered wood are glue-laminated timber and cross-laminated timber. Glue-laminated timber is frequently used in the structural elements of wooden buildings [5].

The primary structural elements of buildings must possess adequate fire resistance to prevent collapse during a fire. Accordingly, Korean construction law mandates fire-resistant structures in buildings exceeding a certain size, requiring that the structural elements maintain performance even when exposed to high temperatures for a specified duration. Structural elements in wooden buildings must also be designed to be fire-resistant. Internationally, char rate and char depth are key concepts in designing fire-resistant wooden structures, with fire resistance performance demonstrated through sections that display the char layer [6]. In South Korea, the criterion for recognizing the fire resistance of wooden elements is based on char depth over time, and wooden building elements must be designed according to the approved char depth. Since the char layer provides minimal load-bearing support, the char properties of wood are crucial for evaluating fire resistance performance [7].

Wooden buildings utilize glue-laminated timber with various cross-sectional sizes depending on the intended use and required fire resistance performance. The cross-section of structural elements influences the usable area and zoning.

The purpose of this study was to conduct fire resistance tests on structural glue-laminated timber beams made from Korean larch (Larix kaempferi) and analyze the char properties. For this purpose, fabricated specimen beams with different cross-sectional shapes and areas underwent load-bearing fire resistance tests. We analyzed the test results, focusing on char depth, char rate, and changes in char thickness based on the aspect ratio of the beams.

2. Wood Charring and Fire Resistance Tests: Theoretical Background

2.1. Pyrolysis and Charring

Wood undergoes pyrolysis within a specific temperature range. As wood is gradually heated from room temperature, the water content begins to evaporate, primarily being driven off around 100 °C. Above this temperature, wood breaks down at the molecular level. Between 100 and 200 °C, carbon dioxide and carbon monoxide are released alongside water vapor in the form of steam. Pyrolysis of wood starts at approximately 120 °C, and at temperatures exceeding 200 °C, charring occurs rapidly due to dehydration.

When heated, pyrolysis results in the formation of a char layer and the production of volatile gases. At temperatures below 300 °C, the char layer forms as a byproduct of pyrolysis. However, at temperatures above 300 °C, the generation of volatile gases increases significantly, becoming the primary product of pyrolysis. After the release of volatile gases, the remaining char layer thickens over time, which tends to reduce the char rate [8].

The boundary between the char layer and the residual section is distinctly defined, with the temperature at this boundary typically around 300 °C [9]. The residual section refers to the structural cross-section that retains the structural performance of wooden elements during a fire. Figure 1 illustrates a cross-section of a glue-laminated timber beam after exposure to high temperatures.

2.2. Fire Resistance Design for Wooden Structures

The formation of a char layer, as described above, protects the residual cross-section, allowing the element to maintain its performance even as the fire progresses. Structural design specifications are based on the residual section, and the final cross-section of the element is calculated by considering the char depth required for the specified fire resistance duration.

South Korea has a system for accrediting the fire resistance of structural elements in buildings. For wooden elements, the char depth verified through fire resistance tests is used to ensure adequate fire resistance performance. In contrast, other countries often use the char depth of the wood, as calculated from the char rate, in fire resistance design.

When wood is exposed to fire, if the temperature exceeds a certain level, burning occurs. When the breadth (b) × length (d) of a section of wood is exposed to fire and reduced to a residual section of b_f × d_f, the char depth of the exposed section can be calculated using the equations below [9].

c = βt

c = char depth (mm)

β = char rate (mm/min)

t = time (min)

As shown in Figure 2, the residual section after the formation of the char layer can be obtained as follows:

b_f (breadth of the residual section) = b − 2c

d_f (length of the residual section) = d − c (heating applied to three faces)

The design char rates recommended in the Eurocodes [6] (European technical standards that offer a harmonized approach to the structural design of buildings) are presented in

Table 1. Design charring rates β₀ and β_n of timber (Eurocode 5).

Material	Minimum Density (kg/m2)	Char Rate
		β0 (mm/min)	βn (mm/min)
Glue-laminated softwood timber Solid or glue-laminated hardwood timber	290	0.65	0.70
	450	0.50	0.55

. The char depth specified in the Eurocode standard is derived from the relationship between the char rate β, char depth, and time, as shown in the equations above. Since corner areas experience approximately 10% more charring compared to the rest of the section, both a one-dimensional char rate (β₀) and a notional char rate (β_n) for corners are provided.

The Wood Handbook published by the American Forest and Paper Association [10] proposes both linear and non-linear models for char depth, while the American Wood Council’s National Design Specification (NDS) for Wood Construction employs the non-linear model. In this context, the species-specific char rates outlined in the ASTM Standard Test Methods for Fire Tests of Building Construction and Materials (E 119) are substituted into the equations below to calculate char depth for use in fire resistance design.

t = mx_c^1.23

t = time (min)

m = char rate (mm/min)

x_c = char depth (mm)

3. Evaluation of Fire Resistance Performance and Results

3.1. Summary of Test Specimens

To investigate the char properties in relation to the aspect ratio and cross-sectional area of glue-laminated timber, we prepared test specimens from Korean larch (L. kaempferi). Fire resistance tests were conducted to examine the influence of aspect ratio on char properties by preparing specimens with the same cross-sectional area but differing widths and heights. The density of the test specimens ranged from 290 to 450 kg/m³.

As major structural elements of buildings, beams are typically load-bearing, necessitating relevant load-bearing fire tests. The observed char properties, including char depth behavior and char rate, were analyzed. Loading was calculated based on the residual section, using the design char rate suggested in Eurocode 5. A 60 min fire test was conducted with a test load corresponding to the calculated residual section (load ratio 1.0).

To evaluate the adequacy of fire resistance performance, we applied the displacement criteria suggested for load-bearing performance in the Korean industrial standard KS F 2257-1 (Methods of fire resistance tests for elements of building construction—general requirements) [11].

Table 2. Summary of fire resistance test specimens.

Specimen	Specimen Size (mm) (W × H × L)	Test Time (min)	Test Load (kN)	Aspect Ratio (W/H)	Moment of Inertia (cm4)
B-L-1	500 × 600 × 4700	60	80.5	5/6	900,000
B-L-2	200 × 600 × 4700		13.1	1/3	360,000
B-L-3	200 × 500 × 4700		10.5	2/5	208,333
B-L-4	200 × 400 × 4700		7.5	1/2	106,667
B-L-5	400 × 200 × 4700		4.7	2/1	26,667

Specimen Name: B (Beam)-L (Load-bearing fire test)-1~5.

characterizes the glue-laminated timber beam specimens of Korean larch prepared for this experiment.

The moment of inertia of structural glue-laminated timber beams varies with the aspect ratio. The beams were observed to experience vertical stress, potentially resulting in differing char properties between the upper and lower sections. This study examined the correlation between the beam’s char performance and moment of inertia in relation to the cross-sectional area ratio.

3.2. Test Summary

The fire resistance performance tests were conducted in accordance with Korean industrial standards KS F 2257-1 and KS F 2257-6 (Methods of fire resistance tests for elements of building construction—specific requirements for beams) [12]. For the fire temperature curve, we utilized the standard time-temperature curve from ISO 834-1 (Figure 3) [13] and measured the char depth of the specimens as well as their displacement under load.

The testing procedure employed a fire resistance test furnace equipped with a burner using liquefied natural gas as fuel. The furnace had three closed sides. A load was applied at four points on the top of the specimen using a hydraulic loading device positioned at the top of the furnace. The peak load was set at 500 kN, with a peak compression length of 1000 mm. Figure 4 illustrates the loading device and the configuration of the load-bearing tests.

3.3. Measuring Char Depth

Char depth was measured from sections taken at three evenly spaced locations along the length (L) of each specimen. At each section, char depth was measured at four locations, resulting in a total of 12 measurements per specimen. Figure 5 illustrates the locations for sectioning each specimen and for measuring char depth at each section.

3.4. Test Results

Table 3. Results of fire resistance tests.

Specimen	Specimen Size (mm) (W × H × L)	Displacement (mm)	Charring Depth (mm)	Charring Rate (mm/min)	Aspect Ratio (W/H)
B-L-1	500 × 600 × 4700	1.9	37.3	0.62	5/6
B-L-2	200 × 600 × 4700	2.2	37.5	0.63	1/3
B-L-3	200 × 500 × 4700	0.6	38.1	0.64	2/5
B-L-4	200 × 400 × 4700	2.0	43.2	0.72	1/2
B-L-5	400 × 200 × 4700	5.1	38.3	0.64	2/1

presents the fire resistance results for the glue-laminated beams made of Korean larch with varying cross-sections. We assessed the performance of the specimens based on displacement under load, following the guidelines outlined in KS F 2257-1 (Methods of fire resistance test for elements of building construction—general requirements). All five specimens exhibited significantly less displacement than the maximum allowable limit, confirming their compliance with fire resistance standards.

With the exception of B-L-4, which showed a considerable charring process, the char depth for the remaining specimens ranged from 37.2 to 38.3 mm.

Table 4. Cross-sections of timber beams after the fire resistance test.

Specimen	B-L-1	B-L-2	B-L-3	B-L-4	B-L-5
Before the fire resistance test
After the fire resistance test

shows photographs taken after the fire resistance performance test, displaying the cross-sections from which the measurements were recorded.

4. Discussion

4.1. Change in Displacement

All five glue-laminated Korean larch beam specimens underwent a 60 min fire resistance test in a furnace. All specimens exhibited very stable displacement behavior, meeting the performance criteria. While it is typical for elements like beams to demonstrate increased displacement with rising load, our test specimens did not show any significant differences in displacement behavior. Figure 6 illustrates the changes in displacement over the test duration.

The minimal change in displacement observed in the glue-laminated timber beams in our study is likely due to differences in the method of calculating the test loads. Typically, test loads for beam components are calculated relative to the total cross-sectional area of the specimen. However, since glue-laminated timber develops a char layer, the load is assessed based on the residual area after the char layer is removed. This char layer continuously protects the structural section throughout the fire resistance test, resulting in minimal damage to the structural integrity. At load ratios where the test load is less than the external forces considered during design, it seems that the loads were insufficient to significantly impact the specimens’ displacement. In this study, the change in displacement did not appear significant. In the future, we aim to conduct a study on the amount of load according to the cross-sectional area by changing the amount of load.

4.2. Changes in Char Depth

The examination of the effects of cross-sectional size on char depth across five specimens. By comparing the estimated char depth based on the Eurocode standard (applying a safety factor of 10% of the load) with the measured char depth of the specimens (excluding specimen B-L-4), found a difference of 3% relative to the one-dimensional char rate (β₀) and 10% compared to the notional char rate (β_n).

In Figure 7, which illustrates the char depth measured at each of the 12 locations for each specimen, variations in char depth measurements were observed between the width and height dimensions due to the heating of the beam elements on three sides. This discrepancy likely explains why only specimen B-L-4, which had a shorter height, exhibited more significant charring.

In this study, the difference in charring depth according to the section area ratio was not significant, except for B-L-4. The charring depth according to the width and height ratios did not change significantly in the width ratios, but there was a difference in the charring depth in the height ratios.

4.2.1. Effects of Width on Char Depth

We compared the char depth between specimens B-L-1 and B-L-2, which were of the same height (H). As shown in Table 3, although these two specimens had different cross-sectional areas due to their varying widths (W), the char depths were comparable.

4.2.2. Effects of Height on Char Depth

We compared the char depths among specimens B-L-2, B-L-3, and B-L-4. As shown in Table 3, the cross-sectional areas varied significantly due to their differing heights (H). While relatively weak, Figure 8 indicates a trend suggesting that char depth develops more rapidly with shorter heights.

4.2.3. Effects of Width (W) and Height (H) on Char Depth in a Constant Cross-Sectional Area

We compared the char depths between specimens B-L-4 and B-L-5. Although these specimens had the same cross-sectional area of 800 cm², they exhibited a 13% difference in char depth. The graph in Figure 9 illustrates the relationship between cross-sectional area and char depth for each of the five specimens. Based on these results, we concluded that when the cross-sectional area remains constant, greater char depth occurs when the width is shorter than the height. In summary, despite identical cross-sectional areas, the differing moment of inertia values indicate a variation in char properties.

5. Conclusions

We conducted load-bearing fire resistance tests on beams made from glue-laminated Korean larch timber and obtained the following results regarding their char properties.

First, with the exception of specimen B-L-4, all other specimens exhibited a char depth that differed by 3% compared to the one-dimensional char rate (β₀) and 10% compared to the notional char rate (β_n) as estimated by the Eurocode standard, applying a safety factor of 10% of the load. We concluded that, for the smaller specimens, the beam’s width influenced char properties more significantly than its length.

Second, in glue-laminated timber, the residual cross-section serves as the structural section that withstands external forces. We determined that applying an external force with a 1.0 load ratio on the residual section had minimal impact on displacement.

Third, for smaller specimens, we observed that at a constant cross-sectional area, charring was more pronounced when the width was less than the height. This phenomenon is attributed to vertical forces acting on the beams, which cause the char layer in the lower part to dislodge and subsequently be replaced.

Given the increasing use of engineered wood products, glue-laminated timber has garnered attention as a viable construction material. In this study, we analyzed the char properties of load-bearing glue-laminated beams made from Korean larch (L. kaempferi). We believe our findings can serve as a valuable reference for research and development in fire resistance design standards for wooden structures utilizing Korean timber, particularly when comparing results with char rates and char depths documented in other markets and jurisdictions.