Abstract
The effects of wood species and panel connections on the vibration and heavy-weight impact sound insulation performance of cross-laminated timber (CLT) slabs were investigated. CLT panels (5-ply, 150 mm thick, 1 m wide, and 4.2 m long) made of larch (Larix kaempferi) and pine (Pinus densiflora) were manufactured with 30 mm thick laminae, considering three types of joints. Three CLT panels of the same species and joint type were connected using spline joints, butt joints, or half-lap joints to form 3 m wide and 4.2 m long slabs for testing. The floor impact sound insulation performance of the CLT slabs was measured according to KS F ISO 10140-3, using the standard heavy-weight impact source, a rubber ball. Additionally, four accelerometers were installed at 400 mm intervals beneath the CLT slabs to analyze the deflections and natural frequencies of the slabs. The results of the experiment indicated that there were no significant differences depending on the wood species and the CLT panel joints. These findings suggest that wood species and joint methods can be flexibly applied in the design of CLT slabs.
Download PDF
Full Article
Fundamental Natural Frequency and Floor Impact Sound Insulation Performance of CLT Slabs Based on Wood Species and Panel Connections: An Experimental Study
Sung-Jun Pang,a Hyo-Jin Lee,b Yeon-Su Ha,c Chul-Ki Kim,b Ho-Jeong Cho,a and Sang-Joon Lee b,*
The effects of wood species and panel connections on the vibration and heavy-weight impact sound insulation performance of cross-laminated timber (CLT) slabs were investigated. CLT panels (5-ply, 150 mm thick, 1 m wide, and 4.2 m long) made of larch (Larix kaempferi) and pine (Pinus densiflora) were manufactured with 30 mm thick laminae, considering three types of joints. Three CLT panels of the same species and joint type were connected using spline joints, butt joints, or half-lap joints to form 3 m wide and 4.2 m long slabs for testing. The floor impact sound insulation performance of the CLT slabs was measured according to KS F ISO 10140-3, using the standard heavy-weight impact source, a rubber ball. Additionally, four accelerometers were installed at 400 mm intervals beneath the CLT slabs to analyze the deflections and natural frequencies of the slabs. The results of the experiment indicated that there were no significant differences depending on the wood species and the CLT panel joints. These findings suggest that wood species and joint methods can be flexibly applied in the design of CLT slabs.
DOI: 10.15376/biores.20.1.100-120
Keywords: Cross-laminated timber (CLT); Vibration; Floor impact sound; Stiffness; Panel joints
Contact information: a: Department of Wood Science and Engineering, Chonnam National University, Gwangju, Republic of Korea; b: National Institute of Forest Science, Seoul, Republic of Korea; c: Korea Institute of Civil Engineering and Building Technology, Goyang-si, Republic of Korea;
* Corresponding author: lsjoon@korea.kr
Timber buildings are becoming taller, with an increasing number of floors (CTBUH 2023). Figure 1 shows a global overview of mass timber buildings, either completed or under construction, specifically those that are eight stories or taller. There is a noticeable trend towards constructing taller and more multi-story buildings, indicating a growing preference for structures exceeding eight stories. The development of cross-laminated timber (CLT) has significantly contributed to the increasing height and number of floors in timber buildings (Duan et al. 2022; Younis and Dodoo 2022; Bhandari et al. 2023). CLT is an engineered wood product that consists of layers of lumber boards stacked and glued together in alternating directions (Balasbaneh and Sher 2021; Santos et al. 2021; Ayanleye et al. 2022). This crosswise layering enhances the material’s strength, stability, and structural performance (Brandner et al. 2016; Siddika et al. 2021; Shulman and Loss 2023).
Korean national standards for the quality of CLT have been developed to produce and use CLT from domestically sourced wood (KS F 2081 2021; Oh et al. 2023; Yang et al. 2023, 2021). The structural performance of Korean CLT made from local wood species has been evaluated (Pang and Jeong 2019, 2018; Pang et al. 2021), along with the shrinkage and expansion characteristics in response to moisture content (Pang and Jeong 2020). Research also led to the development of hybrid CLT using plywood (Choi et al. 2015, 2018; Pang et al. 2019), and hybrid slabs combining CLT and concrete (Oh et al. 2023; Pang et al. 2022; Quang Mai et al. 2018), with their structural characteristics thoroughly evaluated. These studies primarily have focused on structural safety. However, the vibration and noise between floors cause significant anxiety and discomfort for users (Jeon 2001; Jeon et al. 2009; Jo and Jeon 2019). Research on the vibration and noise characteristics of CLT slabs is essential for high-quality residential timber buildings.
Fig. 1. The rise of tall timber buildings (CTBUH 2023)
The issue of inter-floor noise in high-rise apartments is a common concern in densely populated urban areas, not unique to Korea (Gibson et al. 2022; Kang et al. 2023). This issue often arises from factors such as footsteps, moving furniture, loud music, or other activities that generate sound and vibrations, which can easily travel through a building’s structure. Korea enforces the world’s strictest inter-floor noise standards. To construct an apartment with more than 30 units in Korea, the impact sound must meet the standard of 49 dB or less (Ha et al. 2023). These stringent building codes and regulations pose challenges to the expansion of tall timber buildings.
CLT panels are manufactured by laminating solid wood, thereby inheriting the characteristics of the chosen wood species. Additionally, the dimensions of CLT panels are limited by the size of pressing machinery and practical considerations related to transportation. When used as slabs or walls in buildings, CLT panels should be assembled on-site to function as integral components of slabs or walls. This study aimed to investigate whether wood species selection and panel connection methods impact the vibration and noise characteristics of CLT slabs. Furthermore, the correlation between the vibration performance and noise performance of CLT was analyzed, and an attempt was made to estimate noise performance based on the vibration performance of CLT.
MATERIALS AND METHODS
Specimens
Eight types of test specimens were prepared with different wood species, thickness, and CLT panel joints (Table 1). The specimen ID format indicates the wood species with the first letter, the joint type with the second letter, and the CLT thickness with the number, which comes third. The test specimens with CLT panel thicknesses of 300 mm and 450 mm were classified as “mixed” because they designed stacking 150 mm thick CLT panels with different types of CLT panel joints.
The CLT panels were made with two species, larch (Larix kaempferi) and pine (Pinus densiflora), respectively. The dimensions of each lamina were 30 mm in thickness, 130 mm in width, and 4,200 mm in length. The laminas were graded according to KS F 3020 standard using a machine grader (MGFE-251, IIDA Kogyo, Komaki, Japan).
Table 1. CLT Specimens and Experimental Results
The graded laminas were laminated into five layers to achieve CLT grades, C-E10-E8 (Larch CLT) and C-E8-E6 (Pine CLT), as specified in KS F 2081 standard. In the CLT grade nomenclature, the first letter signifies CLT, the second letter denotes the grade of the outermost layer, and the third letter denotes the grade of the inner layer. The grade of lamina used in both outer layers of the CLT is identical, and similarly, the grade of lamina used in the three inner layers is also the same.
Fig. 2. Joint types used to connect the CLT panels
The adhesive used to glue laminas was Phenol – Resorcinol – Formaldehyde (PRF). A quantity of 200 g/m2 was applied for each layer. The press pressure was set at 1 MPa, followed by a 20-hour pressing period and an additional week for curing. The mean moisture content of the CLT was 12 ± 2%, and the specific gravity of the produced CLT panels were 0.587 for larch CLT and 0.476 for pine CLT. The overall dimensions of the CLT panels were 150 mm in thickness, 1000 mm in width, and 4,200 mm in length.
The CLT panels were connected using spline joints, half-lap joints, or butt joints to form 3000 mm in wide and 4200 mm in length for tests. Figure 2 shows the three joint types, all of which utilize self-tapping screws. For the spline joints, Ø 6 mm × 80 mm screws (HBS model, Rothoblass) were used. For the half-lap joints, Ø 6 mm ×130 mm screws (HBS model, Rothoblass) were used. For the butt joints, Ø 7 mm × 180 mm screws (VGZ model, Rothoblass) were installed at a 45° angle.
The vibration characteristics and impact sound insulation performance of the CLT specimens were evaluated following the floor impact sound laboratory measurement methods outlined in KS F ISO 10140-3 (2021). Figure 3 illustrates the setup of a CLT specimen and the designated locations for the impact source. The tests were conducted at an internationally certified testing agency, Fire Insurers Laboratories of Korea, specializing in assessing floor impact sound insulation performance.
Figure 3 shows the internal laboratory environment and experimental setup. A standard heavy-weight impact source, a rubber ball (Rion, YI-01), was dropped from a height of 1 m at five points (Fig. 3 (b)). These positions included the center of the slab and four corners positioned 750 mm away from the borders of the slab (Fig. 3 (c)). Each drop of the rubber ball was repeated nine times at each specified position.
To analyze the deflection and natural frequency of the CLT slab during impact testing, four accelerometers (B&K, type 4527) were installed beneath the CLT slab at 400 mm intervals, as illustrated in Fig. 3 (c) and Fig. 4 (a).
Fig. 3(a & b). Installation of CLT specimen and locations for impact source hitting
Fig. 3 (c). Installation of CLT specimen and locations for impact source hitting
Fig. 4. Installation of accelerometers and microphones
For measuring the floor impact sound insulation performance of the CLT slabs, five microphones (B&K, type4189) were installed at a height of 1.2 m within the sound receiving room (Fig. 4 (b)).
Analysis Method for Vibration and Acoustic Behavior of CLT Slabs
The displacement and natural frequency of CLT slabs were obtained from the recorded acceleration response spectrum. Figure 5 outlines the process and method used to analyze the acceleration spectrum. The displacement of CLT slabs was computed by performing double integration of the acceleration spectrum over the time axis. The natural frequency of the CLT slab was derived by transforming the time domain of the displacement response to the frequency domain using the Fast Fourier Transform (FFT) (He et al. 2023; Xie et al. 2020).
The floor impact sound insulation performance of CLT slabs was evaluated using a single-number quantity base on the KS F ISO 717-2 standard (KS F ISO 717-2 2020). The sound pressure level, recorded using the microphone, was divided into the 1/3 octave band based on the frequency. The evaluation frequency for the floor impact sound was 50 to 630 Hz. The single-number quantity for the heavy-weight floor impact sound (Li, Fmax) was calculated using Eq. (1),
(1)
where m is the number of impact source locations and is the maximum floor impact sound level for impact source location j.
Fig. 5. Derivation of displacement and natural frequency of CLT slab from acceleration data
Stiffness of CLT Slabs by Impact Load
The three types of stiffness related to the behavior of the slab, bending stiffness (EI), dynamic stiffness, and impact stiffness, were determined. Bending stiffness represents the structural member’s resistance to deformation under bending. It is determined by the material’s elastic modulus (E) and the geometric properties of the member, specifically the moment of inertia (I). The EI value was derived by Eq. (2), and more information can be found in the CLT handbook (FP Innovations 2014).
(2)
In Eq. 2, is the effective bending stiffness of CLT (N·mm2), is the modulus of elasticity of ith layer (MPa), is the width of ith layer (mm), hi is the thickness of ith layer (mm), is the cross-section area of ith layer (mm2), and is the distance between the center point of ith layer and the neutral axis (mm)
Dynamic stiffness is associated with the vibration of the slab and indicates how the slab behaves at specific frequencies. Dynamic stiffness was determined using the mass of the slab and the vibration frequency measured by accelerometers. Equation (3) describes the relationship between vibration frequency, stiffness, and mass under the assumption that the structure undergoes simple harmonic motion. Based on Eq. (3), Eq. (4) was derived to calculate the dynamic stiffness.
(3)
(4)
where f is the natural frequency measured by acceleration response spectrum (Hz), kdynamic is the dynamic stiffness of CLT slab (N/m), and is the mass of CLT slab (kg).
The impact stiffness is related to the deflection of the slab due to out-of-plane loads. The impact stiffness of the test slab was calculated by dividing the applied loads by the resulting deformation, as described in Eq. (5). The load acting on the slab comprises the impact force generated by the test ball, the self-weight of the slab, and the weight of the experimenter. The impact load was generated using a hollow silicone rubber ball with a diameter of 185 mm and a thickness of 30 mm. The ball, weighing 2.5 ± 0.2 kg, was dropped from a height of 1 meter, producing an impact load of approximately 1,500 N (Ha et al. 2023; Hyo-Jin Lee et al. 2023; KS F ISO 10140-3 2021; Yazbec et al. 2022),
(5)
where kimpact is the impact stiffness of CLT slab (N/m), Wimpact is the impact load by the rubber ball (1,500 N), WCLT is the self-weight of CLT panel (N), Wexperimenter is the weight of the experimenter (N), and 
is the maximum deflection of CLT slab (mm)
The maximum deflection of the slab was simulated using finite element method (FEM), since it is difficult to experimentally measure the deflection shape and deflection at all points of the slab. The FEM was performed using Midas NFX software (Midas IT 2024), and the input parameters, material properties, are presented in Table 2. The properties of the laminas were assumed to be those of an orthotropic material.
In Table 2, the elasticity for each grade was set to the median value of that grade according to KS F 3020. The elasticity in the transverse direction was defined as 1/30 of the longitudinal elasticity, based on the CLT handbook (FP Innovations 2014). The shear modulus in the longitudinal direction was set to 1/16 of the longitudinal elasticity, and the shear modulus in the transverse direction was set to 1/10 of the longitudinal shear modulus.
Poisson’s ratios were applied based on values from the Wood Handbook (Forest Products Laboratory – USDA 2021). In Poisson’s ratio notation, the first subscript denotes the stress direction, and the second denotes the lateral deformation direction. For stress applied in the longitudinal direction, deformation in the radial direction is very small, and thus, no specific value for 
is provided. Therefore, simulations varied 
between 0.01 and 0.1.
The predicted deflection values were then compared with experimental displacement measurements obtained using accelerometers. The actual deflection of the CLT slab was calculated by double integrating the acceleration spectrum (Fig. 5).
Table 2. Material Properties for Finite Element Analysis
RESULTS AND DISCUSSION
Effects of Wood Species and Panel Connections on Vibration Characteristics
The vibration characteristics of a floor are mainly determined by its natural frequency, which significantly affects its impact sound insulation performance. In this study, impact sound insulation was evaluated according to the KS F ISO 10140-3 standard by measuring responses at five specified impact locations. The effects of wood species and panel connection configurations on the vibration behavior of CLT slabs were systematically analyzed based on these measurements, providing insights into how material and structural factors influence acoustic performance.
Figure 6 shows the impact-induced vibration response spectrum of CLT slabs obtained using the FFT. Each graph for an impact location presents 36 frequency spectra (4 accelerometers × 9 repetitions). The amplitude in the vibration frequency spectrum varied across different experiments and accelerometer positions. However, the dominant frequency, which exhibited the largest amplitude, remained consistent regardless of the accelerometer position and the number of repetitions. This phenomenon was observed in all graphs in Fig. 6, irrespective of the impact locations and wood species.
A slab can exhibit multiple natural frequencies, each corresponding to a different vibration mode, such as bending or twisting. The first mode typically appears at the lowest frequency, with higher modes occurring at successively higher frequencies. In the frequency spectrum, peaks indicate these natural frequencies, each associated with a specific vibration mode. Therefore, the dominant peak in Fig. 6, which had the highest amplitude, can be considered the primary natural frequency that best characterizes the slab’s overall dynamic behavior.
The amplitude of a graph in Fig. 6 refers to the displacement of the CLT specimens at a specific frequency. The frequency with the highest amplitude corresponds to the natural frequency that has the greatest influence on the deflection of the slab. The effect of the dominant frequency on displacement was greatest when the impact occurred in the center for all test specimens. When the impact occurred at the four corners, the effect of the dominant frequency on displacement was similar. This indicates that the largest displacement occurred when the impact took place in the center, and the difference was not significant when the impact occurred at the corners.
Table 3 presents the test results based on wood species, CLT panel joint type, and the thickness of the CLT specimens. The natural frequency of the test specimens ranged from 20 Hz to 22 Hz, exceeding the minimum natural frequency requirement of 8 Hz specified in BS EN 1995-1-1 (Kang et al. 2023) and 9 Hz specified in the CLT Handbook (FP Innovations 2014). Specifically, the natural frequencies of L-M-450 and P-B-150 specimens were 22 Hz, while all other specimens exhibited a natural frequency of 20 Hz. Therefore, there was little difference in natural frequency due to variations in wood species and CLT panel joints.
The vibration performance of CLT is influenced by the boundary conditions, as well as the grade and size of the panel (Breneman et al. 2021; Huang et al. 2020). In this study, the size of the test specimens and the test conditions in the laboratory were consistent. Therefore, the results indicate that the boundary conditions likely governed the natural frequency of the CLT specimens, rather than the wood species (density) of the CLT panels and the structural performance of the CLT panel joints.
Table 3. Experimental Results
Relationship between CLT Slab Stiffness and Floor Impact Sound Insulation Performance
In this study, the three types of stiffness related to the behavior of the slab, namely the bending stiffness (EI), impact stiffness, and dynamic stiffness, were determined. Bending stiffness represents the structural member’s resistance to deformation under bending. It is determined by the material’s elastic modulus (E) and the geometric properties of the member, specifically the moment of inertia (I). The EI value was derived by Eq. (2), and more information can be found in the CLT handbook (FP Innovations 2014).