Acoustic Design of Lightweight Timber Frame Constructions PART 3

ACOUSTIC DESIGN OF LIGHTWEIGHT
TIMBER FRAME CONSTRUCTIONS
Bart Ingelaere, Belgian Building Research Institute, Brussels
[email protected]
CHAPTER 4: APARTMENTS
Report COST Action FP0702 : Chapter 4 Acoustic Design of lightweight timber

frame constructions
Net-Acoustics for Timber based Lightweight Buildings and Elements
Working Group 4: Building acoustics design

Action FP0702
Forests, their Products and Services 3/63
CONTENTS
4 - APARTMENT CONSTRUCTIONS ............................................................ 5

4.1 - GENERAL ........................................................................................... 5
4.2 - PARTY WALLS .................................................................................... 10
4.3 - COMPARTMENT FLOOR CONSTRUCTIONS (INCL. CEILINGS) ................................ 10
4.3.1 - Introduction .......................................................................................... 10
4.3.2 - Impact sound insulation .......................................................................... 10
4.3.3 - Direct airborne sound insulation ............................................................... 34
4.3.4 - Floors and comfort against vibrations ....................................................... 34
4.3.5 - Complete floor systems and their direct airborne and impact insulation:
examples ........................................................................................................ 36
4.4 - JUNCTIONS AND FLANKING TRANSMISSION .................................................. 46
4.4.1 - Techniques used to reduce the flanking transmission .................................. 46
4.4.2 - Quantifying the flanking transmission ....................................................... 51
4.5 - REFERENCES .................................................................................. 61
Action FP0702
Special thanks to Lieven De Geetere, Charlotte Crispin, Debby Wuyts, Manuel Van
Damme, Margo Colson, Victor Desarnaulds, Yves Martin, Christian Mertens and
Jean-Luc Kouyoumji for rereading and writing parts of this text.
We would also like to thank everyone who contributed by sharing information and
in particular the members of the COST action, the companies Machiels Building
Solutions, Knauf, National Research Council Canada (Berndt Zeitler), Lignatur,
Gyproc, Fermacell, Renggli AG and Dataholz.com for all the technical data we
could use.
Last but not least, we would like to thank IWT for the research project
‘optimisation of light weight timber frame constructions AH+’ without which it
would not have been possible to do this work.
This text is intended for acousticians with less experience in lightweight timber
frame construction technology and for people with little background in acoustics
working in the building sector with lightweight timber frame constructions.
Relatively simplified acoustic concepts will be used to explain how things work
acoustically and why the use of some concepts is advised and the use of others
might not be such a good idea.
Disclaimer:
This document illustrates the state of the art solutions for buildings using timber
and wood materials to achieve satisfactory sound insulations based on research
and experience to-date. It is recommended that implementation of the solutions
provided in this document, be checked and/or validated with the installation
specifications of the various materials with respect to other requirements, such as
allowable maximum deformation, moisture stability, etc. Furthermore, in critical
situations or when you need assurance or written documentation that the design
you have chosen to use meets the specific country regulations, we strongly
recommend you engage the services of an acoustical consultant to assist you.
Action FP0702
4 - APARTMENT CONSTRUCTIONS
4.1 - General
For terraced dwellings, the acoustic problems have mainly to do with the horizontal airborne
and impact sound transmission.
Multifamily constructions imply in most cases dwellings on top of each other. Total
structural decoupling is then of course not possible any more. This implies a direct impact
sound transmission path (which we did not have with terraced houses), numerous flanking
transmission paths and greater difficulty in obtaining direct airborne sound insulation in the
vertical direction. For countries using the DnT quantity, the relation with R’ also becomes
less favourable as the term 10.lg(V/3.S) is much less advantageous (V/S≈average height)
than in a horizontal direction (V/S≈average depth perpendicular to the separating wall) for
larger rooms.
Sound insulation in a vertical direction is crucial for the experienced acoustic comfort, but is
unfortunately rather complicated to optimize. The lack of acoustic comfort most complained
about is low frequency impact noise.
Larger lightweight timber frame apartment constructions are a relatively recent

phenomenon. Smaller constructions, of the kind of terraced units with one or two
apartments on top of each other in each unit, are more frequently met. Standard
constructions of this type all over Europe pretty much look structurally alike and are largely
determined by Eurocode 5 structural calculations. Façade finishings, section of joists and
studs and layers of thermal insulation differ, but real acoustic optimisation can only be seen
in more recent projects. Lots of details (floors, walls, façades and even some junctions) and
corresponding acoustic, thermal and fire data can be found in the excellent database
www.dataholz.com and in many different publications (see literature list) such as ‘Robust
Details’, ‘Acoustic performance of party floors and walls in timber framed buidlings’, etc.
(see literature list). Some innovative systems (the use of elastic joints to reduce flanking
transmission, special damping constructions within floors, etc..) will be shown later on.
In many countries, technical building guidelines covering lightweight timber frame

constructions exist, but the acoustic information mostly remains scarce and limited to single
ratings based upon the frequency range down to 100 Hz. Moreover, building guidelines stick
to solutions that comply with building regulations. As these requirements are suitable to
guarantee acoustic comfort for heavy constructions, but not necessarily for lightweight
constructions, there still are quite a lot of problems to be solved and improvements to be
made.
In the figure below, a Finnish construction (Ylojärvi apartments) is presented with an overall
great acoustic performance. In the charts, spectral information of normalized impact sound
levels and apparent sound reduction indices are compared with the average results in heavy
apartment constructions with floating floors (see discussion in chapter 2). Both impact and
Action FP0702
airborne sound insulation of the Finnish construction are more than ‘respectable’ but remain
in the low frequency bands below the traditional heavy constructions.
Outline of this section:
In the next sections, we will first have a quick look in 4.2. at the party wall construction,
being quite similar as to the party walls discussed in the part of this text about terraced
houses. Next, compartment floors are being examined in 4.3. The impact sound insulation is
extensively treated with topics such as:
(1) the choice between resilient floor coverings and floating floor;
(2) some words explaining how floating floors acoustically work and how this can be
different compared to floating floors with heavy floors;
(3) possible errors with the characterisation of the efficiency of floating floors;
(4) current craftmanship errors in the field;
(5) types of floating floors in LWTF construction;
(6) the effectiveness of dry floating floor systems used in LWTF constructions
(7) the necessity of false ceilings
Section 4.3. also gives some information about the airborne sound insulation and some
basic information about comfort against vibrations. A series of solutions / examples with the
acoustic performance closes this chapter.
Section 4.4. takes a closer look at junctions and the flanking transmission that occurs in
these. Techniques to reduce the flanking transmission are being discussed. In the report of
WG 1, a methodology to estimate the flanking transmission has been described. In this
document, measurements give some indication about the importance of the flanking
transmission for some junctions.
Action FP0702
REF.: Sound insulation

structures of beam-to-column
framed wooden apartment
buildings *** Asko Keronen –
Mikko Kylliäinen *** Tampere
University of Technology
(1997) *** ISBN 951-722-
776-0 *** Ylöjärvi apartments,
appendix 3 page 1 (L’n) and
appendix 5 page 1
Topping: Vinyl *** 3x12.5 mm fibre gypsum boards (30.8 kg/m²)*** mineral wool 30
mm (100 kg/m³)
Floor: ribbed slab (thickness 6 cm), joists ca. 30 x 6 cm²
Cavity: mineral wool 100 mm (30 kg/m³)
Ceiling: wooden spacers 5x5 cm² *** resilient channels *** 2x12.5 mm gypsum
boards(ca. 2x9.4 kg/m²)
80.0 f [Hz] 1 2
50 53.4 45.6
70.0 63 56.9 47.5
80 54.1 49.4
60.0 100 51.5 54.2
125 50.3 57.6
50.0 160 52.8 58.5
200 47.6 56.1
L'n [dB]
40.0 250 47.9 52.0

315 45.6 50.1
30.0 400 46.6 48.8

500 48.2 46.6
630 45.0 44.0
20.0
800 38.7 41.9
1000 34.5 39.7
10.0
1250 28.9 37.6
1600 24.0 35.2
0.0
2000 20.0 32.1
f [Hz]
2500 18.2 29.5
3150 16.6 27.0
1 Ln,w +CI,50 47 dB CI,50-2500= 3 dB CI,100-2500= dB Ln,w = 44 dB 4000 0.0 25.1
REFLn,w +CI,50 49 dB CI,50-2500= 2 dB CI,100-2500= 2 dB Ln,w = 47 dB 5000 0.0 22.6
(part of Figure 30)

Action FP0702
80.0 f [Hz] 1 2
50 37.1 43.6
70.0 63 35.7 41.3
80 37.3 39.0
60.0 100 41.2 39.0
125 52.4 39.9
50.0 160 51.7 41.9
200 55.3 44.7
R' [dB]
40.0 250 58.4 46.5

315 60.6 47.5
400 61.2 50.0
30.0
500 63.3 52.1
630 64.1 54.9
20.0
800 67.6 57.0
1000 72.2 59.6
10.0
1250 74.2 61.7
1600 75.0 64.2
0.0
2000 77.3 66.4
1000
1250
1600
2000
2500
3150
4000
5000
100
125
160
200
250
315
400
500
630
800
50
63
80
2500 73.9 68.6

f [Hz]
3150 75.0 69.3
1 R'living= 64 dB C50-5000= -3 dB C100-3150= -2 dB R'w = 67 dB 4000 75.0 70.8
2 R'living= 56 dB C50-5000= -1 dB C100-3150= -2 dB R'w = 57 dB 5000 75.0 68.6
Figure 30: comparison of normalized impact sound levels and apparent sound reduction
indices between an acoustically very well performing Finnish LWTF floor construction and
the average result in heavy apartment constructions with floating floors (see discussion in
section 2).
Action FP0702
Figure 31: floor construction in the first 6 floors high LWTF project in Steinhausen,
Switzerland (MFZ Holzhausen, © Renggli AG, Sursee).
Action FP0702
4.2 - Party walls
In the horizontal direction, we can refer to the discussion of party walls in terraced houses,
at least for small scale buildings. For larger projects, the required horizontal stability under
wind load or earthquake resistance might mean that using the same total separation
construction is just not feasible. But in different projects in Europe, we have seen that this
problem in large-scale projects is often solved by having a rigid concrete or steel core inside
the building containing staircases and lifts (necessary in any case for lifts), although this
increases building time. All horizontal forces of the LWTF construction are then brought to
bear on this steel or concrete core (e.g. Limnologen Växjö).
Another problem can be penthouses whose floor plans can stretch out over several
apartments situated below. No particular details and measurements as a solution for this
are available, though one could imagine a locally elastically coupling of the load-carrying
floors each time at the party walls of the apartments below. The floating floor could then
continue above these party walls so that visually no gap occurs, while acoustically no real
structural coupling occurs between the two constituent walls of the lower party walls.
4.3 - Compartment floor constructions (incl. ceilings)
4.3.1 - Introduction
Before considering the junctions and the problems with the numerous flanking transmission
paths, it is useful to study in detail the direct insulation against airborne and impact sound.
Particularly impact sound, mainly in the low frequencies (drumming sound) can be a major
problem in LWTF constructions.
Most compartment floor systems (separating two apartments) consist of 3 structured

layers: a floating floor or resilient floor covering is built up on top of the load carrying floor
(a combination of joists and boards) and a ceiling mostly made of gypsum boards. A
problem could be the thickness (exceeding standard thicknesses of 30 cm to 35 cm in heavy
weight constructions) and the weight of these floors when really high performances are
required.
4.3.2 - Impact sound insulation
A basic structure without any kind of resilient floor covering or floating floor just will not
offer sufficient acoustic comfort against impact sound.
Using floating floors to reduce impact sound has some additional benefits compared with
resilient floor coverings; this is discussed in point (1) here below.
It is important to understand how floating floors reduce impact sound (paragraph 2) and
how it is correctly characterized to avoid mistakes and to optimize constructions or to look
for innovations. But the choice of kind of floating floor is less easy than for heavy
constructions and design mistakes are quickly made (paragraph 3).
Action FP0702
Next (paragraph 4) we will look at the professional placing of the floating floor so as to
avoid frequently-made errors. As small errors almost entirely eliminate the benefits of the
floating floor, good craftsmanship is absolutely necessary. Finally (paragraph 5), we will
take a closer look at the different families of floating floor concepts and their acoustic
performances.
4.3.2.1 - Reducing impact noise: the choice between resilient floor coverings and
floating floors
Resilient floor coverings such as carpets and laminate floors on elastic underlays are
sometimes used in LWTF constructions. These solutions work out fine in terraced houses
where the obtained impact sound reduction can be sufficient. Applying these in apartment
constructions is also feasible but presents certain disadvantages: solutions with resilient
floor coverings require a much better performance of the rest of the construction of the
floor. In some countries, the necessary impact sound insulation must be attained even
without the resilient floor covering (e.g. Belgium) owing to legal concerns and discussions:
the change of carpets towards parquet or tiling is sometimes considered as an interior
decoration change and the concept of the building should be such that these changes have
no impact on the building physics of the construction. Floating floors are in this case not
only an advantage but a must.
Last but not least, floating floors are very interesting for limiting flanking sound
transmission, an advantage resilient floor coverings do not offer. There exist constructions
with resilient floor coverings showing sufficient impact reduction in the laboratory to allow
one to hope that they will comply with acoustic requirements in situ. But there is the
problem of the flanking transmission ‘Df’1. Without the floating floor, the load-carrying floor
and especially the boards will be excited directly by the tapping machine (or walking
persons…) and this energy will be transmitted to the load-bearing walls below where it will
radiate as impact noise and added to the directly transmitted impact noise (‘Dd’). There are
four of these flanking paths ‘Df’ and that can add up quite a lot of sound. Two of these
flanking paths ‘Df’ can be more important than the remaining ones. Indeed the propagation
of the vibrational energy injected by the tapping machine will be more rapidly attenuated by
distance in the direction perpendicular to the load-bearing joists (at each crossing of a joist
an extra attenuation happens). So using linings in front of the floor-carrying walls could
possibly be of some help (though we have to take into account perverse effects of the
mass-spring-mass-resonances of the linings). But the best way to cope with these flanking
transmissions is to install optimised floating floors.
1
An international convention is to indicate transmission ways using capitals for the start of the flanking path at the
emission room (with ‘D’ indicating the direct separating floor or wall between the two rooms seen from the emission
side, ‘F’ represents a flanking wall in the emission room most of the time perpendicular to the direct separating wall
or floor). Minuscules are used for the end of the flanking path at the reception side. (with ‘d’ indicating the direct
separating floor or wall between the two rooms seen from the reception side, ‘f’ represents a flanking wall in the
reception room most of the time perpendicular to the direct separating wall or floor’).
Action FP0702
4.3.2.2 - How do floating floors reduce impact noise?
Floating floors are mass-spring-mass systems (see figure below). They may reinforce
vibrations at the resonance frequency of the system, but above this frequency, the
transmission of vibrations (and the afterwards radiated impact sound) is ever more reduced
with increasing frequency (see figure below). Good floating floors are designed in such a
way that the resonance frequency is as low as possible (where the sensitivity of the human
ear is lower or inexistent) generating an important reduction of the impact sound in the
greatest part of the audible spectrum.
Figure 32: Shifting the resonance
frequency in the figure from fr,1 to fr,2
reduces considerably the impact noise.
Although we try to avoid formulas in this
WG 4 report, the following simple formula
is very useful to calculate the mass-
spring-mass resonance frequency fr. It
allows for a better understanding of how
the floating floor system works. The
resonance frequency fr resulting from the
system composed of the load bearing
floor with surface mass m”1 [kg/m²], the
spring with dynamic stiffness s [MN/m³]
(normally an elastic interlayer) and the
floating floor with surface mass m”2
(surface mass refers to the mass that
acts per surface unit of the spring), can
be calculated by:
1  1 1  [Hz]
fr  s.  
2  m"1 m" 2 
Other mechanisms such as internal and
surface damping will also influence the
final impact reduction obtained with the
system.
Unfortunately, this model is only so simple and valid for rigid concrete constructions. For
less rigid light weight timber frame floors, the behaviour can be unexpectedly slightly
different. Dynamic impacts on the topping can sometimes be unable to cause the interlayer
to compress but instead cause the direct deformation of the supporting subfloor. As such,
subfloor and topping are not sufficiently decoupled and the resonance frequency can be less
influenced by changing for instance the stiffness of the elastic interlayer. This effect has
been noticed in measurements in the National Research Council Canada2 and in the
measurements by BBRI discussed in chart 10 in this section 4.3.2.
4.3.2.3 - Floating floors to be used in LWTF-construction are often wrongly

characterized, leading to wrong concepts and too much impact noise. Moreover
2
On reducing low frequency impact sound transmission in wood framed construction - Berndt Zeitler, Ivan
Sabourin, Stefan Schoenwald, Erik Wenzke - National Research Council Canada (Inter.noise 2012)
Action FP0702
most floating floors are often less efficient when applied in LWTF constructions
than identical ones used in heavy constructions.
The efficiency of a floating floor system is expressed by Lw (see EN ISO 717-2). One could
describe this quantity as a single rating that expresses the reduction of the impact sound of
a floor system due to the use of the floating floor. Owing to the still dominant heavy way of
building, most products used in floating floors have been characterized for use on heavy
floors of ca. 14 cm thick concrete (EN ISO 10140 parts 3 and 5), though these standards
also permit a characterization for three types of lightweight timber floors. One has to be
careful not to use the Lw (the weighted reduction of impact sound level) determined on the
reference concrete floor in applications with lightweight timber frame constructions! The
correct Lw of the floating floor applied on lightweight timber floors can be quite a bit
smaller. One should always keep in mind that the Lw characterizes the total mass-spring-
mass system (load-carrying floor – resilient interlayer –floating floor) and not the resilient
interlayer alone. So the same elastic interlayer applied on and under different masses and
types of materials will have a totally different efficiency in reducing impact sound! There are
two reasons for this:
Floating floors equally work as a mass-spring (here the elastic interlayer)-mass system,
with the mass of the concrete load-bearing floor being considerably different from the
lightweight constructions. Moreover, typical screed like solutions (6 to 8 cm at a density of
1800 kg/m³) put on top of the resilient layer in heavy constructions are often much heavier
than the classical floating floor types (e.g. boards) installed in LWTF constructions. That
means that the mass-spring-mass resonance frequency for the same elastic interlayer is
much lower for the heavy type of construction than for LWTF construction. The lower the
mass-spring-mass resonance frequency, the better the impact sound insulation will be. This
is therefore a problem from the outset for LWTF constructions.
Figure 33: impact sound on reference floors as specified in EN ISO 10140-5. The
improvement of the impact sound represents the difference between the impact sound
measured directly on the reference floor and measured with the floating floor.
Action FP0702
But there is another reason why it is more difficult to reduce impact sound with a
lightweight basic floor than with the heavy concrete reference floor. When the tapping
machine is positioned on the concrete floor, it generates more sound in the higher
frequencies than in the lower frequencies. For the same tapping machine placed on the bare
wooden floor (without a ceiling finishing), the opposite is true: more sound power is
radiated in the lower frequencies than in the higher frequencies (see figure here above).
The attenuation effect of a mass-spring-mass system increases with the frequency above its
resonance frequency. Even if one had floating floor systems with the same mass-spring-
mass resonance frequencies, the above reasoning explains why the reduction of the impact
sound offered by the floating floor will still be much larger for heavy floors than for LWTF
constructions.
All this means that one has to be very careful with Lw –values proposed in technical
documentation. Most of them were measured on standard concrete floors and show an
efficiency that is way above what can be attained with lightweight wooden floors. There are
two possible options if one has to choose a floating floor system (or for manufacturers to
characterize their product): one could ask manufacturers for the Lw-value measured on the
most suitable type of reference wooden floor described in EN ISO 10140-5 represented in
the figure above; the even better alternative is to measure the impact noise level Ln,w of the
complete floor with its ceiling in a laboratory construction.
4.3.2.4 - Of course floating floors should ‘float’ and execution errors must be
avoided. No hard contacts should link the floating floor to the adjacent walls or to
the load-carrying floor. Even small hard contacts will almost entirely eliminate all
beneficial effects of the floating floor. In general, the same rules apply as for the
placing of floating floors in heavy constructions.
Resilient strips should be placed between the floating floor and the adjacent walls so as to
avoid a hard connection. Where foils are used as an elastic interlayer, this can simply be
done by folding the foils up to the wall (figures 34 ‘a’ and ‘b’). Especially when working with
screeds and concrete floating floors, these resilient border strips should be placed with
extreme care. They should only be cut off after the tiling or the parquet has been placed so
as to avoid any hard contact with the wall through the floor finishing. Architects and work
surveyors should check that the border strips are still visible after placing the floor finishing
and before placing the plinths (see figure 34 ‘e’). Plinths should be fixed to the walls and
make no hard contact with the floating floor. If desired, an elastic joint filling (silicones…)
can be applied between plinths and floors.
Pipes passing through the floating floor should be detached from the floating floor using
again resilient strips around the pipes. Fixations of whatever equipment (radiators, etc.)
should not make any hard bridges between the floating and the basic floor (see figure 34 c).
The surface on which the elastic interlayer is to be placed should be horizontal. If electric
tubes or water pipes are fixed on top of the load-carrying floor and foils or mats are used as
an elastic interlayer, then a levelling layer should be installed so as to provide a flat surface
Action FP0702
for the correct placement of the elastic interlayer. Before placing the foils or mats, the
surface should be cleaned and free of all objects (nails, screws, debris,…).
Elastic interlayers placed as mats should connect well without gaps in between them. Foils
should have sufficient overlap and are ideally taped together (see figure 34 ‘a’). If two
superposed elastic interlayers are used, it is recommended to superpose them in crossed
orientations (see figure 34 ‘a’). All of this is especially important when screed or concrete is
used for the floating floor. Its ‘liquid placing’ is most sensitive to even small gaps as it does
not have the advantage of bridging gaps as boards do. Small perforations are again no
problem for boards, but can be catastrophic for screeds. Work project leaders should pay
attention to all manipulations that could create holes in the elastic interlayer before the
placing of the screed. This latter should be done as soon as the elastic interlayer is in place
and all other actions in between should be avoided (perforations created by ladders,
wheelbarrows, falling objects….). Care should also be taken during the placing of the screed
(shovels!), using elastic foil around the tripod to avoid punctuating the foils (see figure
34‘d’).
When porous mats in glass wool, rock wool, cellulose fibres or similar materials are used as
elastic interlayers, a plastic foil should be placed on top of these materials to avoid the
liquid screed or concrete penetrating inside the pores and producing a hard contact between
the two floors (figure 34 ‘f’).
The correct placing of the floating floor is not only crucial for impact but also for airborne
direct and flanking insulation!
Action FP0702

Figure34: the correct placing of the floating floor is vital. Even small hard contacts will
eliminate all positive effects! If a material with open cells is used as an elastic interlayer
(e.g. mineral wool, see right picture), a plastic foil should be applied before installing the
screed.
4.3.2.5 - Types of floating floor systems used in LWTF constructions
Paragraph (2) explained how floating floors reduce impact sound. The lower the mass-
spring-mass resonance frequency, the better the impact sound reduction due to the floating
floor in general will be (though some internal and surface damping mechanism will also be a
significant parameter). The simple formula that calculates the resonance frequency shows
us two possible strategies to optimize the floating floor for impact sound insulation.
We can try to reduce the dynamic stiffness ‘s’ in the formula. This will indeed reduce the
resonance frequency, but we cannot do this indefinitely: beyond an optimization value of
Action FP0702
this dynamic stiffness practical reasons quickly limit the possibilities of this strategy. First of
all, the static stiffness should be such that the resilient layer is not overly compressed
locally under the influence of furniture or even persons (otherwise the floor wouldn’t be a
horizontal surface anymore), secondly walking on a too resilient floor can give a strange
heaving feeling!
The second strategy is to lower the resonance frequency by adding mass (more boards,
screed), preferably both symmetrically below and on top of the elastic interlayer. Increasing
the weight of only the basic floor or only of the floating floor will soon become inefficient as
the formula for the mass-spring-mass resonance frequency shows. (It is only in heavy
constructions that increasing the weight of the floating floor leads to a lower resonance
frequency and hence a better performance. This is of course due to the considerably higher
surface mass of the load-carrying floor in these constructions, so that 1/m1” becomes
negligible compared to 1/m2” in the resonance frequency formula.)
Good floating floors display an optimization of the surface masses and the dynamic stiffness
of the elastic interlayer. Lots of products exist that serve as elastic interlayers.
Increasing surface mass for the load-carrying floor can be done by adding extra boards, by
using or adding extra heavy boards (fibre cement boards, extra heavy fibre reinforced
gypsum board,…), by using sand fillings between the joists (a typical German technique) or
on top of the boards (National Research Council of Canada), by grit fillings in honey comb
elements on top of the boards (Fermacell), sand or concrete in case elements (Lignatur),
dry concrete blocks in case elements (Lignatur) with optimization of the damping (to limit
drum sound)…
Similar actions can be undertaken to increase the mass of the floating floor itself. Very often
though, a screed of 6 to 8 cm thickness of concrete is used as this is a relatively cheap and
very efficient way to increase the surface mass. Moreover, this also gives the possibility to
install floor heating.
Up to now, we have always been considering that in a section of 1 m² of the floating floor
system that 1 m² of elastic interlayer covers 1m² of the basic load-carrying floor and
supports 1 m² of the floating top floor. Let’s call this SYSTEM 1 -solutions.
By reducing the surface of the spring, we can also increase the total mass per surface of the
spring (m1” and m2”), lowering as such the mass-spring-mass resonance frequency of the
system and improving the impact sound insulation. This can be done by concentrating mass
so that it bears down line- or point-wise on the elastic pad/interlayer, resulting in a lower
resonance frequency and thus to a better performance. The obvious advantage is that with
existing reasonable masses of load-carrying and floating floors, quite low resonance
frequencies can be obtained. In this e-book we will call the line-wise solutions SYSTEM 2-
solutions and the point-wise solutions SYSTEM 3-solutions. The figure below shows some of
these SYSTEM 2- and SYSTEM 3-solutions.
Action FP0702
Figure 35: illustrations of floating floors SYSTEM 2 solutions based upon the principle of
concentrating mass carrying line-wise on an elastic interlayer.
Picture a: Using the extra 100 mm wide board strips (18 mm high) on centre every 400 mm
instead of placing the 2 particle-boards (Spano 12 mm +18 mm) straight on the rock wool
(Rockwool 504, 140 kg/m³) reduces the impact sound by 4 dB (L’nT,w + CI,50-2500) . In order
to ‘robotize’ the prefabrication of the floor elements, the alternative way (picture c) of fixing
the 100 mm wide boards directly on the basic floor (putting the rock wool and boards on
top of these strips) was examined and showed identical gains (which is logical in a mass-
spring-mass system).
Pictures b: Lewis steel plates have a ribbed surface less than 20 mm high and are placed
perpendicular to the joists of the load-carrying floor upon high density mineral wool
(140 kg/m³) fixed itself on top of the joists and boards system. A concrete mortar is poured
on top of these ribbed plates resulting in a thin layer of concrete (normally around
100 kg/m²). The steel ribs are specifically shaped so as to act as reinforcement steel for the
concrete. The ribbed structure perpendicular to the joists channels the load partly line-wise
and partly point-wise (dominant mass at the points of crossing with the joists) onto the
mineral wood. www.reppel.nl
Action FP0702
Figure 36: illustrations of SYSTEM 3-solutions based upon the principle of concentrating
mass carrying point-wise on elastic pads. Pictures c is from from CDM company (pads
and iso-lats) see www.cdm.be. Pictures d is from Granab Subfloorsystems
www.granab.se
4.3.2.6 - Effectiveness of dry floating floor systems used in LWTF constructions
Field measurements with the standard tapping machine were carried out in a two-storey
timber frame mock-up construction. The goal was to compare different dry floating floor
systems on the same reference floor. The mock-up contained a reference timber floor
construction separating two transmission rooms. From top to bottom the basis floor
construction was composed as follows (see figure below): 18 mm particle board, timber
joists (section: 240 mm x 45 mm, centre-to-centre distance: 400 mm), timber battens
(section: 45 mm x 22 mm, centre-to-centre distance: 400 mm), 12.5 mm gypsum boards,
directly screwed on the timber battens. A mineral wool filling (90 mm, 16 kg/m³) was
applied in the cavity between the timber joists.
Action FP0702
Figure 37: LEFT: construction of the reference timber floor. RIGHT: example of a floating
floor on top of the reference floor.
Different (dry) floating floor systems were installed on this reference floor and examined for
their impact sound reduction capacity. The examined flooring complexes consisted of a
resilient layer loaded with one or two flooring boards (see figure RIGHT). Different types of
resilient layers and board materials were tested in this set-up.
For reasons of time and cost savings, more than 40 different samples were tested on a
limited surface, defined by typical board dimensions, e.g. 120 cm x 260 cm, 122 cm x
244 cm. In this phase of the study, only impact sound insulation measurements were made.
For certain high performing complexes, airborne transmission of the radiated impact noise
in the upper room became noticeable in the high frequencies (but without influence on the
single ratings).
Figure 38: setup of the comparative measurements
First we examined the influence of the different board materials. Tests were carried out on
different combinations and types of boards using the same 20 mm mineral wool
(140 kg/m³). The following types were examined: particle boards (720 kg/m², 12 mm and
18 mm), OSB boards (600 kg/m³, 12 mm and 18 mm), wood fibre cement boards
(1250 kg/m³, 18 mm), fibre cement boards (1180 kg/m³, 12 mm) and fibre reinforced
gypsum boards (1140 kg/m³, 2x 10 mm).
Eight different complexes were tested in this way. In terms of L’nT,w the results are situated
between 58 dB and 63 dB, while the surface mass of the top layers varies from 11 kg/m² to
45 kg/m². This indicates that surface mass is not the only influence parameter, and
Action FP0702
certainly internal and surface damping mechanisms need to be taken into account. For this
reason we did not necessarily find worse results for boards with lower surface mass.
Ranking the tested complexes by their surface mass (see figure chart 1 below), we observe
only slightly higher impact noise levels (1 to 2 dB) for simple OSB and particle boards (11 to
13 kg/m², 63 dB) compared to nearly twice as heavy complexes such as an additional
12 mm board (18 to 22 kg/m²) or 18 mm wood fibre cement boards (23 kg/m²). On the
other hand, for the same surface mass (23 kg/m²) we observe a difference of 2 dB between
the 18 mm wood fibre cement board and the double layer of fibre reinforced gypsum board,
in favour of the latter. This indicates clearly the importance of the nature of the board
material. The lowest impact noise level (58 dB) was found for the heaviest complex
(45 kg/m²) being a double layer of 18 mm wood fibre cement board. Though compared to
the double layer of 10 mm fibre reinforced gypsum board, one had to double the surface
mass to obtain a negligible improvement of only 1 dB in terms of L’nT,w.
Of course, in order to maximise the gain using more and/or heavier boards, this mass
should be equally/symmetrically distributed to both masses in the mass-spring-mass
system as we explained earlier. For comparison reasons this was not done here, maintaining
always the same reference floor.
Chart 1: different board types tested on top of a 20 mm thick mineral wool layer
(140 kg/m³)
Staying with one type of material, in this case particle boards and OSB boards, the single
isolated influence of the surface mass could be observed (see chart 2 below). Tests were
carried out on the mineral wool layer loaded with an 18 mm board and with an additional
12 mm board, screwed to the first board. Although important improvements are found
between 1250 Hz and 2500 Hz, hardly any improvement of the low frequent efficiency is
obtained. In terms of L’nT,w the improvements are confined to 1 or 2 dB. So adding boards is
a sure way to improve the impact sound reduction, but not the most efficient one if no
equivalent mass increase is applied to the load-carrying floor.
Action FP0702
90
80
70
60
] 50
B
d
[
T
'Ln 40
reference floor without floating floor ‐‐ 71(‐1) dB
30
20 mm mineral wool + (18 mm + 12 mm) particle board ‐‐ 62(3) dB
20 mm mineral wool + 18 mm particle board ‐‐ 63(2) dB
20
20 mm mineral wool + (18 mm + 12 mm) OSB ‐‐ 61(4) dB
20 mm mineral wool + 18 mm OSB ‐‐ 63(3) dB
10
0
50 63 80 100 125 160 200 250 315 400 500 630 800 1000 1250 1600 2000 2500 3150 4000 5000
f [Hz]
Chart 2: loading effect on 20 mm mineral wool for two different board types
In order to increase the loading effect, an experimental set-up was put into place consisting
of 100 mm wide particle board strips (c-t-c distance 400 mm) screwed underneath the top
layer so as to obtain a SYSTEM 2-construction. An important performance gain was now
observed for the low and mid frequency range (below 1250 Hz, see figure chart 3). This
tells us that combining both measures, extra boards and intermediate strips, permits a
considerable overall improvement of the impact sound insulation. In terms of L’nT,w a gain of
5 dB due to the intermediate strips is found. Compared to the initial single value of 71 dB
for the ‘naked’ floor, a considerably lower impact sound level of 57 dB is now obtained.
Action FP0702
Chart 3: effect of concentrating load by means of wooden strips between boards and
resilient layer
Focusing now on the nature of the resilient layer, tests were carried out on several
materials, classified into eight different ‘material groups’. The following colour codes were
used to indicate them:
 Yellow: mineral wool layers (20 mm) – 140 kg/m³, 100 kg/m³
 Green: rubber flake foams (10, 20, 30, 40 mm) – 120 kg/m³
 Blue: PU flake foams (10, 2x10 mm) – 80 kg/m³, 100 kg/m³
 Red: (multi-layered) PE foam membranes (2x 3.5 mm, 4x 2 mm, 2x 3 mm, 5 mm,
6 mm, 9 mm)
 Brown: resin-bound rubber membranes (corrugated 8/4 mm, 3 mm, 4.5 mm)
 Purple: elastomer pads (30 mm, 50 mm)
 Grey: PU flake foam pads (50 mm)
 Orange: Wood fibre insulation boards (18 mm, 36 mm) – 270 to 250 kg/m³
Almost 40 different resilient layers were tested under a complex of 12 mm and 18 mm

particle boards, in order to compare their effectiveness regarding impact noise. A brief look
at the single value results (L’nT,w), shows rather small differences between the tested
samples, except for the ‘purple’ and ‘grey’ group, containing all the ‘pads-based’ solutions
(SYSTEM 3) (see figure chart 4). Again this indicates that effective solutions have to be
looked for in ‘discrete’ applications, such as strips or pads, optimizing the mass-spring-mass
effect for the floating floor. In this way, values in the range of 50-56 dB are obtained for
L’nT,w, still with a rigidly connected gypsum board ceiling as described above. However, the
PU flake pads solutions were found to be too resilient to be used in practice. For the other,
more traditional resilient layers, impact noise values ranging from 58 to 63 are found.
Action FP0702
Mineral wool sample Rubber flake foam PU flake foam
E foam membrane
resin-bound rubber membrane elastomeric pads in polyester
wool
Sa
mple of PU flake foam pad wood fibre insulation boards
(under particle fibre boards
Chart 4: L’nT,w results for different type of resilient layers combined loaded with a double
layer particle boards (12 mm + 18 mm)
Action FP0702
80
70
60
] 50 no floating floor
B
d
[
T mineral wool
'Ln
40 rubber flake foam
PU flake foam
30 PE‐foam membranes
resin bound rubber
elastomeric pads
20
PU flake foam pads
wood fibre boards
10
0 3 0 0 5 0 0 0 5 0 0 0 0 0 0 0 0 0 0 0 0
5 6 8 0 2 6 0 5 1 0 0 3 0 0 5 0 0 0 5 0 0
1 1 1 2 2 3 4 5 6 8 0 2 6 0 5 1 0 0
1 1 1 2 2 3 4 5
f [Hz]
Chart 5: spectral comparison for different type of resilient layers tested under (12 mm +
18 mm) particle boards
A comparison based on spectral information (see figure chart 5) indicates a mainly low and
mid frequency improvement in the case of the pad solutions (SYSTEM 3 solution).
80
70
60
] 50
B
d
[
T
'Ln
40
30
no floating floor
20 elastomeric pads (type 1) in polyester wool, 50 mm + double particle board ‐‐ 53(4) dB
elastomeric pads (type 2) in polyester wool, 50 mm + double particle board ‐‐ 54(4) dB
10
0 3 0 0 5 0 0 0 5 0 0 0 0 0 0 0 0 0 0 0 0
5 6 8 0 2 6 0 5 1 0 0 3 0 0 5 0 0 0 5 0 0
1 1 1 2 2 3 4 5 6 8 0 2 6 0 5 1 0 0
1 1 1 2 2 3 4 5
f [Hz]
Chart 6: two different types of elastomeric pads tested in polyester wool
Two different types of elastomeric pads (both 50 mm) were tested. When embedded in a
polyester fibre wool layer, suppressing standing waves in the cavity, 53 dB and 54 dB were
reached in terms of single values. It should be noted that in spite of certain other samples
leading to higher gains in the 200-2000 Hz frequency range, the pad solutions remain the
best-scoring solutions due to their effectiveness below 200 Hz (figure chart 6). In this
frequency range, even for the most effective (thick) membrane (SYSTEM 1), the spectral
Action FP0702
values remain in the region of 70 dB, leading to relatively high L’nT,w values in spite of their
effectiveness in the higher frequencies.
80
70
60
] 50
B elastomeric pads in polyester
[d
T elastomeric pads
L'
n
40
30
20
10
0 3 0 0 5 0 0 0 5 0 0 0 0 0 0 0 0 0 0 0 0
5 6 8 0 2 6 0 5 1 0 0 3 0 0 5 0 0 0 5 0 0
1 1 1 2 2 3 4 5 6 8 0 2 6 0 5 1 0 0
1 1 1 2 2 3 4 5
f [Hz]
Chart 7: influence of polyester wool layer as cavity absorption with pad solutions (SYSTEM 3
solution)
When no sound absorbent cavity filling surrounds the elastomeric pads (and steel channels
are used to support the floating floor), a shift of the resonance peaks is observed in the low
frequency region as well as an increase of the impact noise levels in the high frequencies
(cavity standing waves, see figure chart 7). In terms of single values, a loss of 2 to 3 dB is
recorded (L’nT,w = 56 dB) compared to the pads solutions with cavity filling.
Considering again the more traditional resilient layers tested under a 12 mm + 18 mm

complex of particle boards (see figures chart 4 and chart 5), the lowest value (58 dB) was
found with the thickest solutions, 40 mm rubber flake foam. The least effective solutions
(63 dB) turned out to be the thinnest PE foam membrane solutions. Nevertheless, a value
of 60 dB was recorded for a specific 2x3.5 mm PE foam membrane, while comparable
results (single values) are obtained for the 18 to 36 mm thick wood fibre insulation boards
and a 2 dB higher (!) single value was found for the 20 mm thick mineral wool layer (140
kg/m³). In spite of their impressive performances in the high frequencies, the mineral wool
layers do not seem to be well adapted to the relatively small load from the boards resulting
in rather high levels at low frequencies (resonance zone). Comparable results (61-62 dB)
were found for equivalent thicknesses of PU and rubber flake foams.
A second series of samples was tested for impact noise insulation in the above described
mock-up, using commercially available preassembled floating floor systems. The 8 different
samples were examined with different kinds of resilient layers: mineral wool (10 mm), wood
fibre board (10 mm) or felt (9 mm). All systems consist of fibre-reinforced gypsum boards
of different thicknesses: 2x 10 mm, 2x 12.5 mm or 18 mm. Depending on the manufacturer
of the specific system, the nature of the fibres used to reinforce the gypsum boards may
differ (same colour indicates same manufacturer).
Action FP0702
Large deviations are recorded in the high frequency range when comparing systems with
similar top layer but different resilient layer (figure chart 8). Felt and wood fibre board seem
to be less effective sub layers, in favour of the more resilient mineral wool.
80
70
60
] 50
B
[d
T no floating floor
'Ln
40 18 mm gypsum fibre board + 10 mm mineral wool [26 kg/m²]
18 mm gypsum fibre board + 10 mm wood fibre [24 kg/m²]
2x 12.5 mm gypsum fibre board + 9 mm felt [32 kg/m²]
30 2x 10 mm gypsum fibre board + 10 mm mineral wool [26 kg/m²]
2x 10 mm gypsum fibre board + 9 mm felt [26 kg/m²]
2x 10 mm gypsum fibre board + 10 mm wood fibre [26 kg/m²]
20
2x 10 mm gypsum fibre board + 10 mm wood wool [26 kg/m²]
2x 10 mm gypsum fibre board + 10 mm mineral wool [26 kg/m²]
10
0 3 0 0 5 0 0 0 5 0 0 0 0 0 0 0 0 0 0 0 0
5 6 8 0 2 6 0 5 1 0 0 3 0 0 5 0 0 0 5 0 0
1 1 1 2 2 3 4 5 6 8 0 2 6 0 5 1 0 0
1 1 1 2 2 3 4 5
f [Hz]
Chart 8: spectral comparison of different pre-assembled dry floating floor systems on

reference floor
In spite of the large high frequency spectral deviations, the single values differ only slightly
and are situated between 61 dB and 63 dB (figure chart 9). The similar, rather poor
effectiveness of these ‘ready made’ systems in the low frequencies, limits the results in
term of single values.
Action FP0702
72 18 mm gypsum fibre board + 10 mm mineral wool [26 kg/m²]
18 mm gypsum fibre board + 10 mm wood fibre [24 kg/m²]
70 2x 12.5 mm gypsum fibre board + 9 mm felt [32 kg/m²]
68 2x 10 mm gypsum fibre board + 9 mm felt [26 kg/m²]
2x 10 mm gypsum fibre board + 10 mm wood fibre [26 kg/m²]
66 2x 10 mm gypsum fibre board + 10 mm wood wool [26 kg/m²]
] 64
B
d
[ 63
,w
T 62 62 62
'L 62
n
61 61 61 61
60
58
MW WF felt MW felt WF WF MW
56
54
Chart 9: single value results for different pre-assembled dry floating floor systems tested on
reference floor
The limits of the mass-spring-mass model for lightweight timber frame floors can be seen in
chart 10. Reducing the stiffness ‘s’ of the elastic interlayer by increasing its thickness does
not lead towards a downward shift of the resonance frequency. The explanation for this is to
be found in the lack of rigidness of the subfloor and was already mentioned with the
introduction of the mass-spring-mass model. Dynamic impacts on the topping can
sometimes be unable to cause the interlayer to compress but instead cause the direct
deformation of the supporting subfloor. As such, subfloor and topping are not sufficiently
decoupled and the resonance frequency can be less influenced by changing for instance the
stiffness of the elastic interlayer. Creating rigid load-bearing subfloors is not only a good
idea for vibration comfort (see section 4.3.4.), it will also improve the impact sound
insulation with floating floors due to the above mentioned effect.
Action FP0702
Chart 10: influence of the thickness of the elastic interlayer. Increasing the thickness of the
rubber flake foam diminishes its stiffness. Doubling the thickness from 10 mm to 20 mm
reduces the stiffness with 2; 40 mm thick foam only has a quarter of the stiffness of 10 mm
foam. One would expect a downwards shift of the mass-spring-mass resonance with
diminishing stiffness, but this was not noticeable in the measurements.
4.3.2.7 - The necessity of false ceilings
We have seen in paragraph (3) that the same floating floor placed on top of a lightweight
timber floor offers a less efficient reduction than installed upon a heavy (concrete) floor.
This lack of efficiency explains also why impact sound in most solutions cannot be solved
only with a floating floor on top of a light joist/boards system: an additional suspended
ceiling will be almost always necessary.
In single family houses, ceilings are often fixed directly on wooden battens (wood furring
strips) identical to the reference floor in the previous paragraph. This is not such a problem
within the same dwelling, but if the floor separates two apartments (compartment floor),
then this solution might not be such a good idea. Although there is some decoupling by the
wood furring strips fixed perpendicularly to the joists, reducing the structural coupling to
point contacts, too much structural sound transmission still occurs.
The ideal solution is a suspended ceiling that has no structural contacts at all with the load-
carrying floor. This is possible to achieve with metal stud systems (see technical manuals of
manufacturers), but only for limited spans. For spans above 4 m, stud heights of 15 cm are
necessary. As this all comes below the joists, very quickly important floor thicknesses are
the consequence (at least when the joists are perpendicular to the metal profiles). One
possibility for increasing the span is to subdivide the span into two or more smaller spans
using wooden beams (on which the metal studs are fixed just as if it were a wall) that can
be placed between the joists of the load-carrying floor. The alternative is a wooden joist
system from wall to wall and completely independent from the load-bearing floor joists.
Action FP0702
Figure 39: maximum room widths of free-spanning fireboard ceilings are limited (see
documentation www.gyprocplafonds.nl and www.knauf.de )
The alternative of a completely independent ceiling is the use of resilient metal channels
that are fixed directly in the joists of the load-carrying floor (see figures 40 ‘a-f’ below).
Action FP0702
Figure 40: (a) Wood furring strips fixed directly and perpendicular to the joists of the floor,
this is only suitable for floors within the same building. (b) Mounting device to fix any heavy
boards (c) (d) Ceiling metal profiles (Knauf) allowing a resilient connection with the joists
(e) The finishing of the joint between wall and ceiling has an influence on the direct and
flanking insulation. (f) Another type of fixation of metal ceiling profiles allowing for a larger
cavity (Gyproc).
Figure 41: So-called resilient Z-

channels (picture from PrimeWall®
Resilient Channel). Possible mistakes
can deteriorate the acoustic
performance. Using the wrong
screws that are too long and enter
the joists can block the resiliently
hung ceiling (this can also happen
with the channels in the above
picture d). Especially the first screws
fixed through the gypsum board can
push the ‘free end’ of the Z-profile
against the joists. Even well-
dimensioned screws can then enter
in the joists, blocking the resilient
system.
Action FP0702
There are still other alternatives of rigid, ‘punctual’ fixation clips fixed to the channels (in
which finally the gypsum boards are screwed).
Research work in NRCC showed the very beneficial effect of increasing the spacing of
resilient channels from 406 mm to 610 mm. The improvement is quite large and ranges
from 4 to 6 points in all cases (see table below). This large increase occurs, because by
reducing the number of resilient channels, the overall stiffness of the connection decreases,
meaning the resonance frequency shifts downwards also. This means that the improvement
due to adding resilient channels starts earlier.3
Table 1: increasing the spacing between the resilient channels results in large
improvements of the impact sound insulation (data from table 5 of the article referenced to
in the footnote).
Specially developed clips with elastic fixations are also available although the possible gain
in insulation remains more limited than when used with concrete floors (see figures and
tables below).
The impact sound insulation will improve with the surface mass of the ceiling. In practice
this means more (fire resistant) gypsum boards or fire resistant gypsum boards combined
with heavier boards. The added mass will lower the resonance frequency and hence
increase the sound insulation.
Inside the cavity, flexible porous material should be added to avoid standing waves and to
help increase the sound insulation. Taking into account fire requirements, very often rock
wool and cellulose fibres are used. The effect of adding these materials inside the cavity will
increase with thickness only if the decoupling from the load-carrying floor is sufficient (if
not, the structural transmission path will be dominant).
3
On reducing low frequency impact sound transmission in wood framed construction - Berndt Zeitler, Ivan
Sabourin, Stefan Schoenwald, Erik Wenzke - National Research Council Canada (Inter.noise 2012)
Action FP0702
Table 2: direct airborne insulation of different resilient ceiling suspension systems and
profiles. The reference floor is a simple wooden floor made of joists and a single OSB panel
of 18 mm. Although no measurements are available below 100 Hz, as a conclusion one can
say that the difference between all the systems (carrying two gypsum boards of 12.5 mm)
is rather negligible. The last two lines represent the single ratings and the spectrum
adaptation terms for the frequency area between 100-3150 Hz (airborne sound insulation)
and between 100 H-2500 Hz (impact sound).
Where the suspended ceiling touches the walls, a hard connection can arise between the
load-carrying floor and the suspended ceiling. Moreover, extra flanking transmission paths
will occur. So the use of an elastic joint between ceiling and wall is from an acoustic point of
view preferable. Unfortunately problems might arise with fire requirements. In some
countries, the use of an elastic junction is allowed in combination with rock wool in the
cavity (forming an extra fire barrier), but in other countries, this is not the case.
Action FP0702
4.3.3 - Direct airborne sound insulation
The direct airborne sound insulation of floors is very similar to what can be said about party
walls. Due to the height of the floor joists and the need for an independent or resilient fixing
of the false ceiling (creating extra cavity height), large cavities that can be filled up with
flexible porous materials are present, allowing for a very low mass-spring-mass resonance
frequency. For the lower frequencies, this is of course ideal.
In party walls, all structural coupling can be avoided except near the foundations. The same
perfect structural decoupling cannot be attained with floors. So the ultimate performance of
the direct airborne sound insulation will always be slightly influenced by some form of
structural sound transmission. In order to obtain ever better performances, cavity width can
be increased and the surface masses of the composing mass-spring-mass system can be
increased (heavier ceilings, heavier complex of basic floor and floating floor). These
strategies have already been commented on above in terms of further increasing impact
sound insulation.
In well-structured decoupled systems, adding thicker porous flexible materials (rock wool,
cellulose fibres,…) will further increase the direct airborne sound insulation. In general, it is
not the direct airborne sound insulation that causes the major worries. Impact sound,
vibrations and flanking airborne sound transmission are the topics in LWTF apartment
constructions that are most difficult to master
4.3.4 - Floors and comfort against vibrations
Not only impact sound is a worry, also vibrations can be experienced in the same and
adjacent rooms when someone is walking around or when children are playing and jumping
around (cups starting to tremble, …). In accordance with EC5 ‘Serviceability under
vibrations of wooden floors’, an accurate design and calculation of lightweight constructions
such as wooden floors is most important. Calculation aspects and requirements are treated
in the reports of WG2 and WG3.
Lightweight constructions are far more sensitive to vibrations than heavy constructions: For
a given vibratory energy, the amplitude of vibration will increase for the lightweight
structural parts of the building. So for a given induced energy, coming either from normal
users of the floor or from external sources, the vibration velocity will be much greater than
in the case of a normal concrete floor.
The second drawback of wooden floors is the anisotropy coming from the great contrast
between the flexural rigidity in the two directions of the floor. There is a direct mathematical
link between the contrast of rigidity and the number of flexural modes of the floors under
40 Hz. From this number of flexural modes, the accelerance of the floor, which is the ratio
of acceleration and induced force, can directly be deduced. The accelerance expresses a
kind of ‘deformability’ or ‘flexibility’ and is a good parameter for the quantification of
discomfort for users. If the number of flexural modes is low, the floor will be in compliance
with EC5 and users will not experience any kind of vibration inconvenience. For example, in
Action FP0702
the case of an isotropic concrete floor of classical size there is only one mode lower than 40
Hz. A wooden floor which fails to respect the rules of good design will reveal up to 7 modes!
Good rules of design in accordance with EC5 are:
 creating floors that are as rigid as possible (especially reinforcing the flexural rigidity
perpendicular to the joists (diminishing the effects of orthotropic behaviour that
otherwise exists);
 keeping the first mode of vibration as high as possible in the frequency domain in
which vibration energy is induced by normal walking of users. The stipulated minimal
limit in EC5 is 7 Hz;
 calculation is always necessary, given that simple building guidelines are just not
enough.
Vibratory energy from walking, dancing etc. is well known in terms of induced force and in
terms of frequency content. In this way the rules of good design have been established in
Eurocode 5. But there can also be problems with exterior sources of vibrations induced by
traffic (especially near places where speed bumps are installed or in the proximity of
deteriorated road infrastructure). In the case of external sources of vibration, frequency
content and amplitude depend on the environment and possibly cannot be met by the
calculation design of EC 5. So discomfort can be experienced by people, even when the
rules of good design of EC 5 have been respected.
Action FP0702
4.3.5 - Complete floor systems and their direct airborne and impact
insulation: examples
R'w= 67.5 dB
R'living= 57 dB
C50‐5000= ‐10.6 dB
C100‐3150= ‐2.5 dB
Ctr,100‐3150 ‐8.4 dB
Ln,w= 49.4 dB
Ln,w+CI,50‐2500= 57.5 dB
CI,50‐2500= 8.1 dB
CI,100‐2500= 0.4 dB
Ref. SDH 08098‐03‐R (Anlage 9),
Bauphysik Iphofen
Topping: KNAUF BRIO WF (complex of 10 mm wood fibre insulation and 18 mm gypsum fibre board)
Floor: Particle board 22 mm *** joists 180x120 mm² o.c. 625 mm
Cavity: 100 mm mineral wool
Ceiling: channels 60 mmx27 mm fixed with ‘Knauf Direktschwingabhänger’ (vibration isolated fastener, see
above), creating a space of 70 mm between joists and ceiling. *** 1x12.5 mm ‘K diamant board’ (ca. 13 kg/m²)

R'w= 71.0 dB
R'living= 62.2 dB
C50‐5000= ‐8.8 dB
C100‐3150= ‐2.3 dB
Ctr,100‐3150 ‐7.8 dB
Ln,w= 45.0 dB
Ln,w+CI,50‐2500= 51.7 dB
CI,50‐2500= 6.7 dB
CI,100‐2500= ‐0.1 dB
Ref. SDH 08098‐03‐R (Anlage 16),
Bauphysik Iphofen
above), creating a space of 70 mm between joists and ceiling board. *** 2x12.5 mm ‘Knauf diamant board’ (ca.
13 kg/m²)

Action FP0702
R'w= 74.6 dB
R'living= 70.1 dB
C50‐5000= ‐4.5 dB
C100‐3150= ‐2.1 dB
Ctr,100‐3150 ‐6.7 dB
Ln,w= 37.8 dB
Ln,w+CI,50‐2500= 43.9 dB
CI,50‐2500= 6.1 dB
CI,100‐2500= 1.3 dB

Ref. SW 07024‐10R, Bauphysik
Iphofen
Ceiling: completely free hanging ceiling on 2X CW‐75 channels, 25 mm below the joists *** 2x12.5 mm ‘Knauf
diamant board’ (ca. 13 kg/m²)
R'w= not avail.
R'living= not avail.
C50‐5000= not avail.
Ctr,100‐3150 not avail.
Ln,w= 71 dB
Ln,w+CI,50‐2500= 71 dB
CI,50‐2500= 0 dB
CI,100‐2500= 0 dB

Ref. 01 011‐T‐48, Bauphysik Iphofen
Topping: none
Cavity: ca. 100 kg/m² sand (ca. 6 cm) on OSB board
Ceiling: wooden battens 50 x 30 mm² (o.c. 50 cm) rigidly fixed to joists*** 2x12.5 mm ‘Knauf GKB board’ (ca.
720 kg/m³)

R'w= not avail.
Ln,w= 68 dB
Ln,w+CI,50‐2500= 70 dB
CI,50‐2500= 2 dB
CI,100‐2500= 1 dB

Topping: 20 mm fibre gypsum board (Gipsfasern Integral) + 10 mm wood fibre insulation (Steico)
720 kg/m³)

Action FP0702
R'w= not avail.
Ln,w= 62 dB
Ln,w+CI,50‐2500= 63 dB
CI,50‐2500= 1 dB
CI,100‐2500= 0 dB

Topping: none
above), creating a 35 mm between joists and ceiling board. *** 1x12.5 mm ‘GKB board’ (ca. 720 kg/m³)

R'w= not avail.
Ln,w= 55 dB
Ln,w+CI,50‐2500= 60 dB
CI,50‐2500= 5 dB
CI,100‐2500= 3 dB

above), creating 35 mm between joists and ceiling board. *** 1x12.5 mm ‘Knauf GKB board’ (ca. 720 kg/m³)
R'w= not avail.
Ln,w= 49 dB
Ln,w+CI,50‐2500= 56 dB
CI,50‐2500= 7 dB
CI,100‐2500= 1 dB

above), creating a space of 35 mm between joists and ceiling board. *** 2x12.5 mm ‘Knauf GKB board’ (ca.
720 kg/m³)

Action FP0702
R'w= not avail.
Ln,w= 74 dB
Ln,w+CI,50‐2500= 75 dB
CI,50‐2500= 1 dB
CI,100‐2500= 0 dB

Topping: none
Cavity: mineral wool 160 mm 35 kg/m³
720 kg/m³)

R'w= not avail.
Ln,w= 65 dB
Ln,w+CI,50‐2500= 67 dB
CI,50‐2500= 2 dB
CI,100‐2500= 1 dB

Topping: 20 mm fibre gypsum board (Gipsfasern Integral) + 10 mm wood fibre insulation (Steico)
720 kg/m³)

R'w= not avail.
Ln,w= 60 dB
Ln,w+CI,50‐2500= 69 dB
CI,50‐2500= 9 dB
CI,100‐2500= 2 dB

Topping: none
above), creating 30 mm between joists and ceiling board. *** 1x12.5 mm ‘Knauf GKB’ (ca. 720 kg/m³)

Action FP0702
R'w= not avail.
Ln,w= 54 dB
Ln,w+CI,50‐2500= 63 dB
CI,50‐2500= 9 dB
CI,100‐2500= 2 dB


R'w= not avail.
Ln,w= 49 dB
Ln,w+CI,50‐2500= 60 dB
CI,50‐2500= 11 dB
CI,100‐2500= 1 dB


R'w= not avail.
Ln,w= 55 dB
Ln,w+CI,50‐2500= 65 dB
CI,50‐2500= 10 dB
CI,100‐2500= 2 dB

Topping: none

Action FP0702
R'w= not avail.
Ln,w= 38 dB
Ln,w+CI,50‐2500= 48 dB
CI,50‐2500= 10 dB
CI,100‐2500= 0 dB

Cavity: ca. 100 kg/m² sand (ca. 6 cm) on OSB board and mineral wool 60 mm
GKB board’ (ca. 720 kg/m³)

R'w= not avail.
Ln,w= 45 dB
Ln,w+CI,50‐2500= 50 dB
CI,50‐2500= 5 dB
CI,100‐2500= ‐1 dB

Topping: none
Cavity: ca. 100 kg/m² sand (ca. 6 cm) on OSB board and mineral wool 60 mm between channels

R'w= not avail.
Ln,w= 41 dB
Ln,w+CI,50‐2500= 50 dB
CI,50‐2500= 9 dB
CI,100‐2500= 1 dB

Cavity: mineral wool 160 mm and mineral wool 60 mm between channels

Action FP0702
R'w= not avail.
Ln,w= 51 dB
Ln,w+CI,50‐2500= 56 dB
CI,50‐2500= 5 dB
CI,100‐2500= 1 dB

Topping: none

R'w= not avail.
Ln,w= 54 dB
Ln,w+CI,50‐2500= 56 dB
REF.: (1)
LiFavg,Fmax= 59 dB
On reducing low frequency impact sound transmission in wood framed construction ‐
(1)
Berndt Zeitler, Ivan Sabourin, Stefan Schoenwald, Erik Wenzke ‐ National Research Li,Fmax,AW= 57 dB
(1)
Council Canada (Inter.noise 2012) ‐ Ref. NRC‐K01 NRCC Heavy impact Ball measurement

Topping: none
Floor: 2 x 19 mm OSB *** joists ca. 5x 25 cm (2”x10”) o.c. 406 mm
Cavity: mineral wool 150 mm
Ceiling: RC spaced 610 mm o.c.! 3 x 12.5 fire rated gypsum boards.

R'w= not avail.
Ln,w= 43 dB
Ln,w+CI,50‐2500= 44 dB
REF.: (1)
LiFavg,Fmax= 42 dB
On reducing low frequency impact sound transmission in wood framed construction ‐ (1)
(1)

Topping: Prefab concrete slab(100 mm) *** 20 mm closed cell foam resilient (no more information available)

Action FP0702
R'w= not avail.
Ln,w= 44 dB
Ln,w+CI,50‐2500= 44 dB
REF.: (1)
LiFavg,Fmax= 44 dB
(1)

Topping: Prefab concrete slab(70 mm) *** 20 mm closed cell foam resilient (no more information available)

R'w= not avail.
Ln,w= 40 dB
Ln,w+CI,50‐2500= 42 dB
REF.: (1)
LiFavg,Fmax= 41 dB
(1)

Topping: Prefab concrete slab(70 mm) *** 20 mm closed cell foam resilient *** 50 cm sand (ca. 80 kg/m²)

R'w= not avail.
Ln,w= 43 dB
Ln,w+CI,50‐2500= 44 dB
REF.: (1)
LiFavg,Fmax= 42 dB
(1)

Topping: Prefab concrete slab(70 mm) *** 20 mm closed cell foam resilient (no more information available) *** 50
cm sand (ca. 80 kg/m²)

Action FP0702
R'w= not avail.
Ln,w= 47 dB
Ln,w+CI,50‐2500= 50 dB
(1)
REF.: LiFavg,Fmax= 46 dB
Berndt Zeitler, Ivan Sabourin, Stefan Schoenwald, Erik Wenzke ‐ National Research
Li,Fmax,AW= 49 dB
(1)


R'w= not avail.
Ln,w= 47 dB
Ln,w+CI,50‐2500= 48 dB
(1)
REF.: LiFavg,Fmax= 46 dB
Berndt Zeitler, Ivan Sabourin, Stefan Schoenwald, Erik Wenzke ‐ National Research
Li,Fmax,AW= 48 dB
(1)

Action FP0702

Figure: the Swiss company ‘Lignatur’ has some specific solutions for
compartment floors. Box and surface elements constitute the load‐bearing
floor. Different types of toppings allow attaining a wide variety of acoustic
performances. The picture down left shows a highly damped solution with
dry concrete blocks and grit fillings, optimizing the low frequency sound
insulation and acoustic comfort (damping the modal peaks and resonances).
Source: http://www.lignatur.ch/2011/en/planning/workbook/
Action FP0702
4.4 - Junctions and flanking transmission
4.4.1 - Techniques used to reduce the flanking transmission
In the optimized acoustic concept for party walls of a continuous cavity from foundations to
roofs, no flanking transmission between horizontally adjacent apartments is possible. But
for apartments one above the other, flanking transmission is definitely present and will limit
the overall airborne sound insulation. The main questions are: ‘how important is this
flanking transmission (see section 4.4.2) and how can we reduce it?’
Figure 42 a: flanking transmission paths exist vertically in all junctions, so also through
room dividing walls of an apartment (left) or via the façades. In the horizontal direction, the
flanking transmission to the adjacent apartment can be eliminated by a party wall such as
described in section 3.
1) The most obvious technique to reduce the flanking transmission is to make a

disconnection / vibration interruption
We have seen this technique being applied to its full extent in the party wall (see above
figure 42 ‘a’), reducing flanking transmission to the adjacent apartment almost totally.
But also smaller disconnections like the use of an elastic joint between the ceiling boards
and the walls (see figure 42 ‘b’) will reduce the flanking transmission from the walls and
floor from the apartment above to the ceiling and walls below. Caution: we were told that
there can be problems with fire safety acceptance in several European countries, making it
necessary to still have a rigid joint.
If the sound insulation is important between two rooms, boards should never continue from
one room to the adjacent room to avoid flanking transmission. This is illustrated in figure
42 ‘c’ with an evaluation of the flanking transmission through a metal stud wall with gypsum
boards/
Action FP0702
Figure 42 b: creating an elastic joint between the ceiling and the walls can reduce the
flanking transmission ‘Fd’ and ‘Df’, but can create problems with the fire safety
requirements in some countries.
TEST 1: REFERENCE TEST 2: REF+4 dB TEST 3: REF+8 dB TEST 4: REF+8 dB TEST 5= REF.+7 dB
TEST 6: REF+9 dB TEST 7: REF +10 dB

Figure 42 c: evolution of the flanking sound insulation for different constructions. The basic
construction was a T junction. The flanking wall was composed of a single layer gypsum
board (12.5 mm) on a single Metal frame (75 mm thick). The cavity was empty in the
reference setup (Test 1). In test 2, the cavity was filled up with mineral wool. In test 3, the
inner leaf was interrupted. The same was done with the outer leaf in test 4. In the tests 5,
6 and 7, the same interventions were made but on a flanking wall consisting of 2x12.5 mm
gypsum boards.
Action FP0702
2) Linings will have some effect. We already discussed the effect floating floors can have
on the 4 flanking transmission paths ‘Df’. Technical linings before the walls in the
emission and the reception rooms will have some effect if fixed with resilient bars,
preferably perpendicular to the wooden studs (see figure 42 ‘d’). Unfortunately no
measurements are known to us to quantify this effect. The empty cavity (necessary to
allow the passage of electric wiring or piping) and the limited width of the cavity will
unfortunately limit the possible benefits, especially in the low frequency bands.
Figure 42 d: technical linings using resilient studs fixed perpendicular to the wood studs can
also diminish some flanking transmission. We do not dispose of any measurements
quantifying this, but we expect the improvement to be only in the mid and higher
frequencies.
3) Using more ‘wood mass’ in the junction apparently also has some effect. In Canada a
‘heavy’ junction with a concentration of wooden beams, showed some improvement
even in the low frequencies (see figure 42 ‘e’). Unfortunately no measurement data is
available that isolates this aspect from other influences. So this hypothesis still has to
be verified with a dedicated setup.
Figure 42 e: creating heavier junctions apparently reduces the flanking transmission.

Unfortunately no measurement data is available to verify this statement (information from
Zeitler Berndt, National Research Council Canada).
Action FP0702
4) Apparently several research groups and consulting offices have tried to use elastic
interlayers to reduce the flanking transmission. Two ‘families of solutions’ can be
seen: the first uses continuous linear elastic interlayers on top of walls and below
floors (or just only below the floors and not interrupting the walls in the project
Limnologen in Växjö, Sweden, figure 42 ‘g’), a second solution uses discontinuous
fixations on top of the load bearing floors (figure 42 ‘h’). This last solution is
apparently only possible with cross laminated timber, the load pressure with
punctual charges being too high for the wood fibres in the horizontal beams of
timber frame constructions. A pragmatic research (figure 42 ‘f’) showed only a small
improvement above 200 Hz that even became negligible when screws were fixed
every 40 cm (necessary to take on the horizontal forces within the construction). The
inefficient behaviour in the low frequencies can be explained by a too small
disruption for the long structural wavelengths of low frequency bending and
transversal waves. In the construction in Limnologen (figure 42 ‘g’), the linear elastic
interlayer seems to have a beneficial effect on the flanking impact sound.
Unfortunately, no measurement results or additional information about this was
communicated.
Figure 42 f: a pragmatic research examined the effect on the airborne standardised sound
insulation DnT of different linear and continuous elastic interlayers on the walls just below
the floor of the room above. The system proved to be only effective above 200 Hz, showing
no difference at all for the low frequencies.
Action FP0702
Figure 42 g: a huge multi-storey lightweight timber frame apartment building in Limnologen

Växjö, Sweden, also uses linear continuous elastic interlayers to reduce flanking impact
sound transmission. All floors bear on the purple elastic interlayer.
Figure 42 h: the use of elastic pads, creating only discontinuous elastic contact points every
150 cm appears to have a better effect in the low frequencies (BBRI-La Maison Idéale
Project).
Action FP0702
4.4.2 - Quantifying the flanking transmission
(1) Example 1: quantifying the flanking insulation of some lightweight timber frame
junctions in a laboratory setup with and without elastic joints.
Methods to predict the total sound transmission (and the insulation against it) have been
developed. A more detailed description can be found in the report of WG 1.
In this chapter some measurements results are given that quantify this flanking
transmission (and vice versa the flanking sound insulation).
Special setups have been built in the BBRI’s laboratories allowing the following
measurements: R, σ, Rij, δ, Ts, Dnf. The measurement of the Dn,f was carried out with the
intensity technique. The sound generated in the source room (C1) was steady and had a
continuous spectrum in the considered frequency range. The radiated power by the
reception wall was measured with an intensity probe. In order to avoid the background
noise from the acoustic hall, we had to use a semi-anechoic box which disrupted the values
below 350 Hz. The Dnf and Rij were then obtained by the following formulas:
and
Figure 43: LEFT: laboratory setup; MIDDLE and RIGHT: a semi-anechoic box (protecting
against the noise influence from the acoustic hall) was used to measure the radiated sound
of the wall via the intensity technique
The measurement procedures and the result analysis are reported in other documents. This
report here summarizes the results measured on (1) lightweight timber frame constructions
in the laboratory, (2) on a research mock-up (built to comply with the EOTA-testing
procedure) and (3) on cross laminated timber constructions.
The first setup to get an idea of the importance of the flanking transmission, was built in the
(former) acoustic laboratory facility in Limelette. Figures 44 ‘a’ to ‘c’ first give a description
of the separating floor and the acoustic performances going from its basic load-bearing
construction (figure 44 ‘a’) to the finished construction in figure 44 ‘c’. Next (figure 44 ‘d’),
a wall is constructed upon this floor creating a L-junction for which the flanking sound
insulations have been determined. In figure 44 ‘e’, the construction has been extended to a
T-junction with a wall on top and below the floor, rigidly connected to the load-bearing
Action FP0702
construction and the ceiling. In figure 44 ‘f’, the effect of an elastic interlayer (reducing the
linear rigid contact to punctual rigid contacts) on the flanking transmission is examined.
Reference floor 3750 x 2400 mm²:
OSB 22 mm (13 kg/m²) *** joists
165x65 mm² , 40 cm o.c.

Topping: /

Rw (C;Ctr)=28 (‐1;‐2) dB
Ln,w (CI)=92(‐5) dB

Ref. BBRI – OSABOIS TEST0
Figure a: load carrying floor
Floor 3750 x 2400 mm²:
165x65 mm², 40 cm o.c.

Topping: honeycomb boards filled
with gravel (Fermacell) 45 kg/m²***
Fermacell boards 2E32 26 kg/m²
(complex of 10 mm MW and 20 mm
gypsum fibre board)

Rw (C;Ctr)=52 (‐2;‐8) dB
Ln,w (CI)=59(1) dB

Figure b: with topping of honeycomb boards filled with gravel Ref. BBRI –OSSABOIS TEST1
Floor 3750 x 2400 mm²:
165x65 mm², 40 cm o.c.
Topping: honeycomb boards filled
with gravel (Fermacell) 45 kg/m²***
Fermacell boards 2E32 26 kg/m²
(complex of 10 mm MW and 20 mm
gypsum fibre board)
Ceiling: totally independent MS
channels,10 mm Fermacell board (7.7
kg/m²), total cavity width 235 mm
filled with 150 mm MW (32 kg/m³)

Figure c: with an independent ceiling (1 board of 10 mm Fermacell) Rw (C;Ctr)=68 (‐3;‐9) dB
(intensity measurement, side ceiling)
Ln,w (CI)=43(1) dB
Ref. BBRI –OSSABOIS TEST2

Figure 44: Floors a, b and c

Action FP0702

Figure d: L‐junction with the studwall rigidly fixed on top of
the floor

Floor 3750 x 2400 mm²:

OSB 22 mm (11.7 kg/m²) *** joists 165x65 mm²,
Rw,Dd = Rw (C;Ctr)=68 (‐3;‐9) dB
40 cm o.c.
(direct sound insulation, measured in the
Topping ‘d’: honeycomb boards filled with gravel
situation corresponding with figure ‘c’, BBRI –
(Fermacell) 45 kg/m²*** Fermacell boards 2E32
OSSABOIS TEST2)
26 kg/m² (complex of 10 mm MW and 20 mm
Rw,Df > 73 dB (flanking sound insulation for the
gypsum fibre board)
transmission from the ceiling ‘D’ to the flanking
Ceiling ‘D’: totally independent MS channels,10
wall ‘f’, determined via intensity measurement
mm Fermacell board (7.7 kg/m²), total cavity
on the gypsum board side of the ‘f’ wall, Ref.
width 235 mm filled with 150 mm MW (32
BBRI –OSSABOIS TEST4)
kg/m³)
Wall ‘f’ 12.5 mm gypsum board, studs 90x40
mm² 60 cm o.c., MW 90 mm, OSB 18 mm

Figure 44: junction d

Action FP0702
Figure e: T‐junction with the studwall rigidly fixed on top
of and below the floor

Rw,Dd = Rw (C;Ctr)=68 (‐3;‐9) dB Floor 3750 x 2400 mm²:
(direct sound insulation, measured in the OSB 22 mm (11.7 kg/m²) *** joists 165x65 mm²
situation corresponding with figure ‘c’, BBRI – 40 cm o.c.
OSSABOIS TEST2) Topping ‘D’: honeycomb boards filled with
Rw,Fd> 76 dB (flanking insulation for the flanking gravel (Fermacell) 45 kg/m²*** Fermacell boards
transmission from the flanking wall ‘F’ to the 2E32 26 kg/m² (complex of 10 mm MW and 20
ceiling ‘d’, value obtained by the energetic mm gypsum fibre board)
subtraction of the value determined by the Ceiling ‘d’: totally independent MS channels,10
intensity measurement result on the ceiling mm Fermacell board (7.7 kg/m²), total cavity
minus Rw,Dd. Ref. BBRI –OSSABOIS TEST11) width 235 mm filled with 150 mm MW (32
Rw,L1L2 > 69 dB (flanking insulation for the flanking kg/m³)
transmission from the flanking wall ‘F’ to the Wall ‘F’ and ‘f’: 12.5 mm gypsum board, studs
flanking wall ‘f’, value obtained by the intensity 90x40 mm² 60 cm o.c., MW 90 mm, OSB 18 mm
measurement result on the wall ‘f’, we supposed
that the shielding of the floor by its topping is
effective enough not to take in account the
transmission path ‘Df’. Ref. BBRI –OSSABOIS
TEST11)

Figure 44: junction e

Action FP0702
Floor 3750 x 2400 mm²:
OSB 22 mm (11.7 kg/m²) *** joists 165x65 mm²,
40 cm o.c
Topping ‘D’: honeycomb boards filled with
gravel (Fermacell) 45 kg/m²*** Fermacell
boards 2E32 26 kg/m² (complex of 10 mm MW
and 20 mm gypsum fibre board)
Ceiling ‘d’: totally independent MS channels,10
mm Fermacell board (7.7 kg/m²), total cavity
width 235 mm filled with 150 mm MW (32
kg/m³)
Wall ‘F’ and ‘f’: 12.5 mm gypsum board, studs
90x40 mm² 60 cm o.c., MW 90 mm, OSB 18 mm

Rw,Dd = Rw (C;Ctr)=68 (‐3;‐9) dB
(direct sound transmission, measured in the
situation corresponding with figure ‘c’, BBRI –
OSSABOIS TEST2)
Rw,Fd> 78 dB (flanking insulation for the flanking
transmission from the flanking wall F to the
ceiling d, value obtained by the energetic
Figure f: similar as in figure ‘e’ but with an elastic junction: subtraction of the value determined by the
T‐junction with the stud wall rigidly fixed below the floor. intensity measurement result on the ceiling
The stud wall on top of the floor stands on an elastic
minus Rw,Dd. Ref. BBRI –OSSABOIS TEST06)
interlayer (CDM 41010, see red strip in drawing). A screw
fixes the top wall every 60 cm through the elastic interlayer Rw,Ff > 71 dB (flanking insulation for the
to the floor, resulting in a punctual rigid connection. The transmission from the flanking wall ‘F’ to the
effect of the resilient interlayer appears to be negligible. flanking wall ‘f’, value obtained by the intensity
measurement result on the wall ‘f’, we
supposed that the shielding of the floor by its
topping is effective enough not to take in
account the transmission path ‘Df’. Ref. BBRI –
OSSABOIS TEST06)

Figure 44 : junction f
(2) Example 2: quantifying the flanking insulation of some lightweight timber frame
junctions in an experimental mock-up (Beringen, MBS).
A major lightweight timber frame manufacturer wants to expand its activities to apartment
constructions. They agreed to cooperate in a research project aiming to generate building
guidelines for lightweight timber frame apartment constructions. A mock-up has been built
that complies with the setup instructions as stipulated by EOTA (see figure a). One of the
experiments was to study the flanking sound transmission in a simple 3-cell timber frame
mock-up and to confront the measurements with the prediction methods as developed by
WG1 (results are discussed in a research report)
Action FP0702
Floor
• Particle board 18
• Joists 240x45 o.c. 400
+ Mineral wool 90
• Wood furring strips o.c. 400
• Gypsum board 12.5 (screwed)
“Exterior” wall
• Particle board 12 (int.)
• Wood studs 140x45 o.c. 400
+ Mineral wool 140
• Softboard 18 (ext.)
“Interior” & partition wall

• Fibre reinforced gypsum board 12.5
• Wood studs 95x45 o.c. 400
+ Mineral wool 95
• Fibre reinforced gypsum board 12.5
Figure 45 a: Timber frame mock-up at Machiels Building Solutions at Beringen (Belgium)

(all dimensions in mm)
Both vibration reduction indices Dv,ij,n using structural excitation and flanking sound
reduction indices Rij using airborne excitation have been measured. In the further analysis,
we will focus only on the flanking sound reduction indices. These have been measured using
the sound intensity technique (using both 50 mm and 12 mm microphone spacers) with
appropriate shielding (see figure b). Only the vertical sound transmission is studied along 2
junctions: the cross junction and the T-junction involving the “exterior” wall. The
measurement results are displayed in figures c and d. Since the direct sound reduction
index - estimated at Rw(C;Ctr;C50-5000;Ctr,50-5000) = 40(-1;-3;0;-6) dB - is much lower than all
the flanking sound reduction indices measured(see figure e), it is clear that, in this case
(very basic floor construction), flanking is not important. However, when the floor
construction is improved considerably, flanking sound transmission may become important.
Action FP0702
Wij
Rij  10lgij  10lg
Ws
i Ws
Wij
j
Figure 45 b: Test setup example for measuring the flanking sound reduction index
100
90
A
80
70
C B
60
Rij [dB]
50 A
40
30 C B
R ij,w (C;C tr;C 50-5000;C tr,50-5000) = 74(-3;-9;-3;-12) dB
20
R ij,w (C;C tr;C 50-5000;C tr,50-5000) = 73(-2;-8;-5;-17) dB
10
R ij,w (C;C tr;C 50-5000;C tr,50-5000) = 62(-2;-6;-2;-12) dB
A
0
125
200
315
500
800
1250
2000
3150
5000
50
80
C B
Figure 45 c: Measured flanking sound reduction indices on the cross junction for the 3 paths
Ff, Fd and Df. Values indicated with a triangle are minimal values, whereas values indicated
with a circle are interpolated or extrapolated values (-5 dB/octave going down at low
frequencies).
Action FP0702
Figure 45 d: Measured flanking sound reduction indices on the T junction for the 3 paths Ff,
Fd and Df. Values indicated with a triangle are minimal values, whereas values indicated
with a circle are interpolated or extrapolated values (-5 dB/octave going down at low
frequencies).
100
90 R'
80
R dir. (simulated)
70
Rij,Cross ‐ D‐f
R, R', Rij [dB]
60
50 Rij,Cross ‐ F‐f
40
Rij,Cross ‐ F‐d
30
20 Rij,T ‐ D‐f
10 Rij,T ‐ F‐f
0
Rij,T ‐ F‐d
50
80
125
200
315
500
800
1250
2000
3150
5000
Frequency band [Hz]
Figure e: Comparison of the (simulated) direct sound reduction index R, the measured
overall sound reduction index R’ and the measured flanking sound reduction indices Rij.
Action FP0702
(3) Example 3: Quantifying the flanking insulation of some cross laminated timber junctions
with and without elastic joints in a laboratory setup
Floor ‘d’ cross laminated timber 9.4 cm thick,

with wooden ‘ribs’ (see drawing) of 9.5x20 cm²,
interdistance 25 cm. The space between the
‘ribs’ is filled with gravel (1400 kg/m³).
Topping ‘D’: 9 mm thick elastic latex lining

Isopack, with a Fermacell dryfloor (fibre
gypsum) system ‘Maxifloor’ of 38 mm thickness
(1100 kg/m³ or 41.8 kg/m²)
Wall ‘F’ and ‘f’: 9.4 cm cross laminated timber

9.4 cm thick, connected with 3 steel connecting
hooks (see picture left) for each wall.
Floor element alone
Rw,Dd= Rw (C;Ctr)=65 (-3;-9) dB
(direct sound transmission through the floor

alone, BBRI –AC5126)
Ln,w (CI)= 50 (-1) dB (BBRI-AC5068)
Flanking sound insulation
Rw, DdFd=61(-2;-8) dB (total insulation for the

transmission paths Dd and Fd determined by
intensity measurement result on the ceiling
minus Rw,Fd. Ref. BBRI –AC5127)
Rw,Ff = 61 dB (determined by intensity

measurement on f, we supposed that the
shielding of the floor by its topping is effective
enough not to take in account the transmission
Figure a: T-junction with cross laminated path Df. Ref. BBRI –AC5127)
timber (RE GT Wal)
Action FP0702
Floor ‘d’ cross laminated timber 9.4 cm

thick, with wooden ‘ribs’ (see drawing) of
9.5x20 cm², interdistance 25 cm. The
space between the ‘ribs’ is filled with gravel
(1400 kg/m³).
Topping ‘D’: 9 mm thick elastic latex lining

Isopack, with a Fermacell dryfloor (fibre
gypsum) system ‘Maxifloor’ of 38 mm
thickness (1100 kg/m³ or 41.8 kg/m²)
Wall ‘F’ and ‘f’: 9.4 cm cross laminated

timber 9.4 cm thick, connected with 3 steel
connecting hooks (see picture left) for each
wall.
Floor element alone
Rw,Dd= Rw (C;Ctr)=65 (-3;-9) dB
(direct sound transmission through the floor

alone, BBRI –AC5126)
Ln,w (CI)= 50 (-1) dB (BBRI-AC5068)
Flanking sound insulation
Rw, DdFd=60(-3;-8) dB (total insulation

for the transmission paths Dd and Fd
determined by intensity measurement
result on the ceiling minus Rw,Fd. Ref. BBRI
–AC5128)
Rw,Ff = 60 dB (determined by intensity

measurement on f, we supposed that the
shielding of the floor by its topping is
effective enough not to take in account the
transmission path Df. Ref. BBRI –AC5128)
Action FP0702
4.5 - REFERENCES
[1] Birgit Rasmussen, ‘Sound insulation between dwellings, Requirements in building

regulations in Europe’, Applied Acoustics, no° 71-2010, pages 373-385.
[2] SBR (several authors), ‘Handboek houtskeletbouw’, 2012, SBR Rotterdam, The
Netherlands
[3] J.D. Quirt and T.R.T. Nightingale, ‘Airborne Sound Insulation in Multi-Family Buildings’,
Construction Technology Update N° 66, 2008, National Research Council Canada
[4] B. Zeitler; I. Sabourin, S. Schoenwald, E. Wenzke, ‘On reducing low frequency impact
sound transmission in wood framed constructions’, Inter.noise 2012, New York City,
US
[5] J.D. Quirt, T.R.T. Nightingale, F. King, ‘Guide for Sound Insulation in Wood Frame
Constructions’, 2006, National Research Council Canada
[6] Berndt Zeitler; Stefan Schoenwald; Trevor Nightingale, ‘Parametric study of sound
transmission through lightweight floors’, Inter.noise 2010, Lisbon, Portugal
[7] Trevor Nightingale, Berndt Zeitler, Stefan Schoenwald, Frances King, ‘A hierarchy of
flanking transmission paths in lightweight wood frame construction’, Inter.noise 2009,
Ottawa, Canada
[8] Van de Kuilen J.W.G., Van der Linden, ‘Timber Frame Housing in Europe’, TNO-report
97-CON-R0763\1, 1997, The Netherlands
[9] ISOVER Multi-Comfort House, Isover Saint-Gobain Benelux, Gent, Belgium
[10] Lignatur Workbook, http://www.lignatur.ch/en/planning/workbook/workbook/ ,

website, 2012, Switserland
[11] Kolb Josef, ‘Bois systèmes constructifs’, Presses polytechniques et universitaires

romandes, 2010, Switserland
[12] Wolfgang Winter, Helmut Schöberl, Thomas Bednar, ‘Holzbauweisen im verdichteten

Wohnungsbau’, Fraunhofer IRB Verlag, 2005, Stutgart, Germany
[13] Barbara Keyworth, Huel Twist, Robin Lancashire, ‘Timber frame construction’, Trada
Technology, 2008, Buckinghamshire, United Kingdom
[14] Beiblatt 1 zu DIN 4109, ‘Schallschutz im Hochbau, Beiblatt 1 zu DIN 4109,

Ausführungsbeispiele und Rechnungsverfahren’, 1989, Berlin, Germany
[15] Asko Keronen, Mikko Kylliäinen, ‘Sound insulating structures of beam-to-column

framed wooden apartment buildings’, Tampere University of Technology, 1997,
Tampere, Finland
[16] G. Pitts, B. Arch (Hons), M. Phil, ‘Acoustic Performance of party floors and walls in
timber framed buildings’, Trada Technology Report 1/2000, 2000, Buckinghamshire,
United Kingdom
Action FP0702
[17] Östman B. and several authors (21), ‘Acoustics in wooden buildings, State of the art
2008, Vinnova project 2007-01653, SP Rapport 2008:16, 2008, Stockholm, Sweden
[18] KNAUF Holztafelbau-Wände W55, 2010, Germany
[19] KNAUF Mehrgeschossiger Holbau Gebäudeklasse 4, 2011, Germany
[20] KNAUF, several test reports, see references in text
[21] Thomas Thode, ‘Leistungen im Holzbau’, Knauf Baukongress 2007, 2007, Germany
[22] Website DATAHOLZ.COM, http://www.dataholz.at/ , Association of the Austrian Wood

Industry, 2012, Vienna, Austria
[23] Delphine Bard, Victor Desarnaulds, Hervé Lissek, Robert Beffa,’ Swiss pendulous
hammer for decoupling measurement of Swiss methodology for decoupling
measurement of service equipment in buildings’, Inter.noise 2012, New York City, US
[24] Fermacell Xella, ‘Constructies voor wanden, plafonds en vloeren’, 2008, Waregem,
Belgium
[25] GYPROC Saint-Gobain, website http://www.gyproc.be/ ,2012, Belgium
[26] Jean-Luc Kouyoumji, Interaction Acoustique et Thermique pour les systèmes

constructifs bois. Projet « Bois AcouTherm », FCBA Bordeaux, 2011, France
[27] Jean-Luc Kouyoumji ; Sylvain Gagnon, ‘Experimental approach on sound transmission

loss of Cross Laminated Timber floors for building’, Inter.noise 2010, Lisbon, Portugal
[28] Anders Ågren, Acoustic highlights in Nordic light weight building tradition - focus on
on-going development in Sweden’, BNAM 2010, Bergen, Norway
[29] Krister Larssona, ‘Conclusions from a state of the art report on acoustics in wooden
houses’, EURONOISE 2009 2008, Edinburgh, Scotland (UK)
[30] Heinz J. Ferk, Sound insulation evaluation results of lightweight and massive built
multi-family houses in Austria, power point presentation, 2009, TU Graz, Austria
[31] D. Wuyts, C. Crispin, B. Ingelaere and M. Van Damme. Laboratory Sound Insulation
Measurements of Improved Timber Floor Constructions: A Parametric Survey.
JOURNAL OF BUILDING ACOUSTICS Volume 13 • Number 4 • 2006
[32] M Van Damme, M Blasco, C Crispin, P Huart, B Ingelaere and D Wuyts. Acoustic
Comfort In Timber Frame Dwellings. BUILDING ACOUSTICS • Volume 14 • Number 1
• 2007 Pages 15 – 24
[33] C. Crispin, B. Ingelaere, D. Wuyts, ‘Détails de construction pour le confort acoustique

dans les habitations II (Relation entre les prestations acoustiques des bâtiments et
celles des éléments qui les composent).’, Rapport MEZ/MAE 2006-2008., 2008, BBRI
Brussels, Belgium
Action FP0702
[34] C. Crispin., ‘Détails de construction pour améliorer l’isolation aux bruits aériens des
constructions à ossature bois multiétagées. Rapport OSSABOIS’, 2010, BBRI Brussels,
Belgium
[35] Lieven De Geetere, ‘Experimental Investigations on Flanking Sound Transmission in

Lightweight Timber Frame Constructions.’,Euronoise 2012, Prague, Chechia

Acoustic Design of Lightweight Timber Frame Constructions PART 3

Uploaded by

Copyright:

Available Formats

Acoustic Design of Lightweight Timber Frame Constructions PART 3

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Acoustic Design of Lightweight Timber Frame Constructions PART 3

Uploaded by

Copyright:

Available Formats

ACOUSTIC DESIGN OF LIGHTWEIGHT

TIMBER FRAME CONSTRUCTIONS

Bart Ingelaere, Belgian Building Research Institute, Brussels

Report COST Action FP0702 : Chapter 4 Acoustic Design of lightweight timber

Net-Acoustics for Timber based Lightweight Buildings and Elements

Working Group 4: Building acoustics design

4 - APARTMENT CONSTRUCTIONS ............................................................ 5

Larger lightweight timber frame apartment constructions are a relatively recent

In many countries, technical building guidelines covering lightweight timber frame

Outline of this section:

REF.: Sound insulation

Floor: ribbed slab (thickness 6 cm), joists ca. 30 x 6 cm²

Cavity: mineral wool 100 mm (30 kg/m³)

40.0 250 47.9 52.0

30.0 400 46.6 48.8

(part of Figure 30)

40.0 250 58.4 46.5

2500 73.9 68.6

4.2 - Party walls

4.3 - Compartment floor constructions (incl. ceilings)

Most compartment floor systems (separating two apartments) consist of 3 structured

4.3.2 - Impact sound insulation

4.3.2.2 - How do floating floors reduce impact noise?

4.3.2.3 - Floating floors to be used in LWTF-construction are often wrongly

4.3.2.5 - Types of floating floor systems used in LWTF constructions

4.3.2.6 - Effectiveness of dry floating floor systems used in LWTF constructions

Figure 38: setup of the comparative measurements

Almost 40 different resilient layers were tested under a complex of 12 mm and 18 mm

Mineral wool sample Rubber flake foam PU flake foam

mple of PU flake foam pad wood fibre insulation boards

(under particle fibre boards

Chart 6: two different types of elastomeric pads tested in polyester wool

Considering again the more traditional resilient layers tested under a 12 mm + 18 mm

Chart 8: spectral comparison of different pre-assembled dry floating floor systems on

4.3.2.7 - The necessity of false ceilings

Figure 41: So-called resilient Z-

4.3.3 - Direct airborne sound insulation

4.3.4 - Floors and comfort against vibrations

Good rules of design in accordance with EC5 are:

4.4 - Junctions and flanking transmission

4.4.1 - Techniques used to reduce the flanking transmission

1) The most obvious technique to reduce the flanking transmission is to make a

TEST 1: REFERENCE TEST 2: REF+4 dB TEST 3: REF+8 dB TEST 4: REF+8 dB TEST 5= REF.+7 dB

Figure 42 e: creating heavier junctions apparently reduces the flanking transmission.

Figure 42 g: a huge multi-storey lightweight timber frame apartment building in Limnologen

4.4.2 - Quantifying the flanking transmission

Figure 44: Floors a, b and c

Figure 44: junction d

Figure 44: junction e

“Interior” & partition wall

Figure 45 a: Timber frame mock-up at Machiels Building Solutions at Beringen (Belgium)

Floor ‘d’ cross laminated timber 9.4 cm thick,

Topping ‘D’: 9 mm thick elastic latex lining

Wall ‘F’ and ‘f’: 9.4 cm cross laminated timber

Floor element alone

Rw,Dd= Rw (C;Ctr)=65 (-3;-9) dB

(direct sound transmission through the floor

Ln,w (CI)= 50 (-1) dB (BBRI-AC5068)

Flanking sound insulation