Proceedings of the Institute of Acoustics

 

 

Comparison of the direct sound insulation for wooden joist, CLT and timber hollow box floors

 

Anders Homb¹, SINTEF Community, Trondheim, Norway

Simone Conta², SINTEF Community, Trondheim, Norway

 

ABSTRACT

 

Timber buildings are now present worldwide, and their market is growing rapidly, pushed by sustainability and environmental arguments. Several technological options for engineered wood systems are available and well developed, the most common being wooden joist, CLT and timber hollow box. In this paper, we present a comparative study of impact sound insulation of the three options. The analysis is based on data collected from laboratory measurements. The results include the comparison of the bare floor construction and examples of assemblies with additional floating floor, suspended ceiling or a combination of the two. We identify the characteristic trends and highlight the effect of selected parameters as for instance mass per unit area, assembly thickness and properties of the resilient layers. The comparison attempts to show the constructive consequences of the different options on the sound insulation properties. It could help selecting the most suitable floor system in future projects and contribute making timber construction more efficient and affordable.

 

1. INTRODUCTION

 

In the process of designing building, planners and architects have to consider a number of aspects from different disciplines in order to choose the most appropriate design and solution. Due to the increasing interest on environmentally friendly raw material, we see a huge interest the last decade in developing floor constructions based on wooden material. Besides traditional wooden joist floors, we notice a rapid development of CLT based solutions, and an increasing interest in timber hollow box floors. Several research projects and measurement programs were established and carried out in recent years making available a large amount of measurement data on wooden floor constructions. But in this "jungle" of possibilities, how should architects and constructors choose the most appropriate solution depending on the boundary condition in each project? Among several parameters influencing this choice we will mention:

• The total floor mass(affect the load bearing system, foundation dimensioning and span length)
• The thickness of the floor construction (affect possible limitations on the number of storeys and integration of installations).
• The carbon footprint level (the floor itself or for the overall building).
• Span length (requirements depending on the building category and architectural plan).
• Fire resistance (affect especially the ceiling solution and use of non-wooden materials)
• Pros and cons due to horizontal sound transmission and vertical flanking transmission.

 

In this paper, we make a first attempt of comparing wooden joist, CLT and timber hollow box floors based on two input parameters, total floor mass and floor thickness, and evaluating the impact sound insulation properties. A comparison including all the parameters mentioned above would be interesting, but it is very demanding and it is far beyond the scope of this work. Qualitative comments will be given regarding span length and limitations due to flanking transmission. The impact sound level data used in this comparison are measured in the laboratory, which is the most accurate possibility of comparing solutions.

 

2. COLLECTION OF IMPACT SOUND MEASUREMENT DATA

 

The data collected for this comparison originate from measurements carried out according to ISO 10140-2 [1]standard or previous versions valid at the time of measurement. The different laboratories in which these measurements were carried out were accredited laboratories following the ISO 10140- 1 [2]. Unfortunately, as indicated in ISO 12999-1 [3], there are no results available of frequency dependent uncertainties for impact sound insulation measurements in laboratory conditions. Inter laboratory measurements (round-robin tests) for impact sound insulation are not very common and especially not sufficient in numbers to evaluate laboratory measurement uncertainty. Standard uncertainties for single-number values of laboratory measurement are however given in ISO 12999-1 standard and estimated to be about 1.5 dB.

 

All measurement results presented in this paper were available in the frequency range from 50 Hz. Selected examples of normalised impact sound pressure levels in the frequency range 50 – 5000 Hz are presented in chapter 3 for representative configurations. From the test result, different single number quantities for rating the impact sound insulation were calculated according to EN-ISO 717-2 [4]; Ln,w, the spectrum adaptation term, CI,50-2500 and the sum of these, Ln,w+CI,50-2500. It can be interesting to compare these values with current requirements, for which we refer to [7].

 

Looking into the different floor constructions, we recognize that some relevant combinations are missing. Measurement data from other floor constructions may exist, but not inside our database so far. When we try to compare different solutions, it is also obvious that supplementary constructions should be on board. Therefore, for some examples the comparison obviously has weaknesses.

 

2.1. Overview Table

 

Among the datasets available, we extracted the most comparable and typical solutions. We then grouped them according to the type of the bare floor (Joist, CLT or hollow box) and to additional layers. The categories are presented in table 1 where we also refer to the corresponding table summa rising the main properties of each configuration considered. We selected here the total mass per unit area, the total thickness, Ln,w-value, the sum of Ln,w + CI,50-2500 as key parameters. The reference to the data source [5], [6], [7], [8] and [9] is also provided for each configuration. Similar data is also collected for other categories, but these data have not yet been processed and analysed.

 

Table 1: Grouping of measurement data presented in this paper.

 

 

2.2. Bare Element

 

Table 2 presents collected data from bare floors. Regarding the Joist and CLT results, given numbers are an average of measurement data from almost equal constructions measured in two different laboratories. Note that the data from the joist solution include an ordinary plasterboard directly attached to the bottom of the floor construction and absorbing material in the cavity.

 

Table 2: Key data for bare floor solutions

 

 

Bare floors are in practice never used without additional layers. Often it is reliable to calculate improvement of the sound insulation when sufficient data of bare floor solution is given. Therefore, information given in table 2 and the frequency spectra in chapter 3.1 will be helpful for calculating alternative floor constructions.

 

2.3. Separate Ceiling

 

Table 3 presents collected data from solutions with an additional, separate ceiling mounted under the floor. Note that floor coverings are missing, i.e. the tapping machine excites directly the CLT or chipboard material.

 

The solutions considered have almost the same mass per unit area and not too different total thickness. They deliver relatively equal impact sound insulation. By adding some top floor construction, both solutions can fulfil the requirements for impact sound insulation in apartment buildings in Germany, Norway, Sweden and Switzerland [7].

 

Table 3: Key data for solutions with a separate ceiling below the bare structure

 

 

2.4 Lightweight Top Floor, Point Elastic Solution

 

Table 4 present collected data from solutions with a lightweight top floor on point elastic units. For all objects, there is no ceiling below the CLT/Hollow box.

 

Table 4: Key data for solutions with lightweight top floor on point elastic units

 

 

The constructions presented here have very different floor thickness. It is nevertheless interesting to compare CLT and hollow box based on their mass per unit area. We see that Ln,w + CI,50-2500 is considerably higher with a hollow box compared to CLT despite comparable mass per unit area, but these hollow box objects is developed for 9 m span length. Two of the solutions fulfil impact sound insulation recommendation for apartment buildings.

 

2.5. Lightweight Hybrid Top Floor, High Hynamic Stiffness

 

Table 5 present data from hybrid floor solutions with a lightweight top floor on continuously elastic layer. For all objects, the dynamic stiffness of the resilient layer is s' > 20 MN/m³.

 

Table 5: Key data for hybrid floor solutions but a lightweight top floor on a resilient layer

 

 

Numbers presented in table 5 show large spreading of all key parameters but note the lack of ceiling for the CLT and Hollow box solutions. As seen from the CLT results, large difference of the impact sound insulation is obtained even if the mass per unit area and floor thickness is at relatively equal level. The same effect is not seen for the presented Hollow box solutions. Some of the construction examples fulfil impact sound insulation recommendations for apartment buildings.

 

3. FREQUENCY SPRECTRUM EXAMPLES

 

Examples of frequency spectrum from laboratory measurements is presented in chapter 3.1 to 3.4.

 

3.1. Bare Floor Element

 

Figure 1 show laboratory measurement results of bare wooden floor constructions. The figure cover solutions/data given in Table 2.

 

Note that the data from the joist solution include an ordinary plasterboard directly attached to the bottom part of the beam and some absorbing material in the cavity. In major part of the frequency range, the impact level difference between the objects is huge. As expected, the lightweight joist floor shows high levels at low frequencies and low levels at medium and high frequencies. The reason for the latter one is an effect of favourable sound radiation properties of the plasterboard in the ceiling. At lower and medium frequencies, the CLT and empty Hollow box solution offer approximately the same impact level, but for the CLT example by significant higher volume of wood. Note also that these hollow box objects is developed for 9 m span length. Hollow box solution including gravel in the cavity offer low impact level at low and medium frequencies. Both softness at top of the element and the sound radiation play an important role at higher frequencies.

 

 

Figure 1: Impact sound level spectrum results from selected Joist, CLT and Hollow box solutions

 

3.2. Separate Ceiling

 

Figure 2 show frequency spectra from available results with an additional, separate ceiling. The figure cover solutions/data given in Table 3.

 

 

Figure 2: Impact sound level spectrum data from available Joist and CLT solutions

 

Both examples show very comparable frequency spectra regarding the impact sound level and the single number quantities is also within the same range due to small differences also at low frequencies. The CLT solution offer the lowest single number quantity, but on the other hand the highest floor thickness.

 

3.3. Lightweight Top Floor, Point Elastic Solution

 

Figure 3 show frequency spectra from solutions covering a lightweight top floor on point elastic units. The figure cover solutions/data given in Table 4.

 

The CLT examples offer low impact levels at medium and high frequencies compared to the Hollow box examples. At low frequencies, the properties depend very much on a combination of element stiffness and dynamic properties of the elastic unit. Regarding the Hollow box examples, note the significant effect of the gravel in the cavity, which contribute to low impact levels at low frequencies and just slightly lower single number quantity as the most preferable CLT example.

 

 

Figure 3: Impact sound level spectrum data from selected CLT and Hollow box solutions

 

3.4. Lightweight Hybrid Top Floor, High Dynamic Stiffness

 

Figure 4 show frequency spectra from hybrid floor solutions with a lightweight top floor on continuously elastic layer. For all objects, the dynamic stiffness of the resilient layer is s' > 20 MN/m³. The figure cover solutions/data given in Table 5.

 

 

Figure 4: Impact sound level spectrum data from selected Joist, CLT and Hollow box solutions

 

Measurement examples given in figure 4 show the huge spreading of the impact sound insulation properties. In general, the limitation is still at low frequencies even for these examples with increased mass. Note also the huge difference between results in the medium frequency range for the CLT examples. Among these examples, the joist solution offers the lowest single number quality, but as mentioned before, this construction example includes a resilient ceiling. At higher frequencies, the Hollow box examples show relatively high impact levels, but it relies on the absence of a floor covering. When installing some type of floor covering, the impact levels at high frequencies is expected to not limit the single number quantity.

 

4. CONCEPT TO COMPARE WOODEN FLOOR SOLUTIONS

 

4.1. Comparison Suggestion

 

The upcoming question now is how we should compare different floor systems. As mentioned in the introduction, several parameters will be relevant for the choice in the programming of new buildings. Necessary span length for instance will depend on the building category and architectural plan. For the future, the carbon footprint will be more important as a mandatory parameter to document in each case. But reliable and neutral numbers on this is still challenging and at least resource demanding to establish. Such parameter is therefore not available for floor constructions presented in this paper, but some examples are given in [10]. The comparison between solutions in this paper will therefor rely on the total mass of the floor, the total height beside the impact sound insulation level itself.

 

Comparison parameter based on the total mass is suggested by the relative mass (RM), determined by the following equation:

 

 

Comparison parameter based on the total floor height is suggested by the relative height (RH), determined by the following equation:

 

The logarithmic expression is selected because of the dB representation of the sound insulation properties. Since better impact sound insulation follow decreasing dB numbers we have to invert the Ln,w+CI,50-2500 by the following equation.

 

 

From this we calculate the suggested evaluation number based on RM: Mass comparison figure = RM/IIL and similar for the suggested evaluation based on RH: Height comparison figure = RH/IIL. The outcome from this is a grading of the floor constructions, respectively based on the total mass and the total height of the constructions. Figure 5 show this grading based on the relative mass based on floor constructions from table 3, 4 and 5 and figure 6 show similar grading based on the relative height.

 

Low number on the Y-axis in figure 5 means preferable impact sound insulation with respect to the mass of the floor. The comparison presented in figure 5 show that a separate ceiling offers the most preferable grading, and the difference between the CLT and Joist solution is marginal. For lightweight solutions including point elastic support, the CLT solutions offer the most preferable properties, but higher floor mass is needed in comparison to the separate ceiling solutions. Increased mass is also necessary regarding the hybrid solutions from table 5 relative to the single number quantity. Grouping in this way, the Joist solution is more preferable compared to both CLT and Hollow box solutions.

 

Low number on the Y-axis in figure 6 means preferable impact sound insulation with respect to the total thickness of the floor. The comparison presented in figure 6 show that hybrid floor constructions may be preferable with respect to floor thickness, especially looking at data from Hollow box solutions. On the opposite, the Wood box examples from table 4 (lightweight top floor, point elastic solution, 9 m span length) show the highest ratio between the relative height and impact sound insulation properties. A general comment is also that the CLT examples shows better or approximately equal grading compared to the joist solutions.

 

 

Figure 5: Grading of floor constructions based on the relative mass. Input from table 3, 4 and 5.

 

 

Figure 6: Grading of floor constructions based on the relative height. Input from table 3, 4 and 5.

 

4.2. Other Aspects to Consider

 

Beyond the parameter that we considered in this paper, there are several other aspects that must be included when comparing the different options. Among them we highlight:

• When long span floorsis necessary; Hollow box orJoistsolutions are far more effective compared to CLT due to the serviceability state requirements
• When horizontal sound transmission is an issue and vertical separation between elements not possible; an additional plasterboard layer is necessary on CLT and hollow box solutions while it was already included in the joist solutions considered in this paper.

 

5. SUMMARY AND CONCLUSION

 

The concept of grading presented in this paper is a preliminary attempt to compare properties. Further development of the concept based on comments and discussions of this work is of course welcome. Still, some preliminary conclusions may be drawn from the given data set, based on the grouping from table 1.

 

Grading with respect to the floor mass:

• Joist solutions, perform significant better compared to CLT and Hollow box solutions
• CLT solutions perform better or equal compared to Hollow box solutions.

 

Grading with respect to the floor thickness:

• Hollow box solutions, perform significant better compared to CLT for the hybrid solution
• CLT solutions perform better grading or equal grading compared to Joist solutions.
• For other comparison of Joist or Hollow box solutions it is necessary also to consider the floor thickness depending on the span length

 

An extract of available measurement data was analysed, but within the evaluation process we also recognize a lack of relevant combinations of ceiling, subfloor and top floor solutions. In this paper, an attempt has been made to evaluate the effect from respectively the mass of the floor and the total thickness on the impact sound insulation properties for different wooden floor solutions. The comparison has been based on the suggested relative mass or relative hight, resting on a chosen log number and reference value for respectively mass and floor height. This is to be discussed, and other numbers may change the relative differences between the objects. Further work is necessary to improve this toolbox for comparing floor constructions.

 

5. ACKNOWLEDGEMENTS

 

Major part of measurement data used for the analysis in this paper are from previous research projects. We will especially mention the Silent Timber Build project, the Woodsol project and RFF Hybrid project for the opportunities this has provided for this work.

 

6. REFERENCES

 

1. ISO 10140-2:2010. Acoustics. Laboratory measurement of sound insulation of building element. Part 3: Measurement of impact sound insulation.
2. ISO 10140-1. Laboratory measurement of sound insulation of building element. Part 1: Application rules for specific products.
3. ISO 12999-1. Acoustics – Determination and application of measurement uncertainties in building acoustics. Part 1: Sound insulation
4. EN-ISO 717-2. EN-ISO 717-2:2013. Acoustics. Rating of sound insulation in buildings and of building elements. Part 2: Impact sound insulation.
5. Homb, A. Guigou-Carter, C., Hagberg, K., Schmid, H. Impact sound insulation of wooden joist constructions: Collection of laboratory measurements and trend analysis. Building Acoustics 2016, Vol. 23(2), 73-91.
6. Homb, A., Guigou-Carter, C., Rabold, A. Impact sound insulation of cross-laminated timber/ massive wood floor constructions: Collection of laboratory measurements and result evaluation. Building Acoustics 2017, Vol. 24(1), 35-52.
7. Homb, A., Conta, S., Kumer, N. Sound insulation of timber hollow box floors. Collection of laboratory measurements and trend analysis. Building Acoustics 2021, Vol. 28(2), 161-183.
8. NS 8175:2012. Acoustic conditions in buildings. Sound classification of various types of buildings. Standard Norge 2012.
9. Homb, A. Hybrid joist floor constructions. Evaluation of measurement results. Proceedings ICA 2019. 23^rd International Congress on Acoustics, Aachen, Germany, 9 to 13. September 2019.
10. Skaar, C., Solem, B. & Rüther, P. Composite floors in urban buildings: Options for a low carbon building design, in: 6Th Forum Wood Building Nordic Trondheim, Trondheim, Norway, 2017.

 

---

¹ anders.homb@sintef.no

² simone.conta@sintef.no