BUILDING DESIGN USING COLD FORMED
STEEL SECTIONS: ACOUSTIC INSULATION

Sound levels and sound insulation values are expressed in decibels (dB), whilst pitch or frequency is expressed in Hertz (Hz).

In the case of sound levels, the dB rating is a representation of the volume of the sound whilst in the case of sound insulation values it is a measure of the amount by which sound transmitted from one room to another is reduced by the separating construction. Some typical sound levels and sound insulation values are shown in Figure 1.

The sound insulation properties of walls or floors vary with frequency and, as most sounds are a mixture of several different frequencies, certain frequencies within a sound are likely to be attenuated more effectively than others by a given construction (low pitched sounds are normally attenuated less than high pitched sounds). In view of this, the sound reduction characteristics of walls and floors are usually measured at a number of different frequencies across the hearing range. The normal frequency range of measurements is shown in Figure 2.

Figure 2: The frequency range of building acoustics measurements

Sound insulation can be described in a variety of ways. This can initially be confusing for the architect or designer trying to interpret specifications or manufacturers' literature. The following attempts to explain some of the main terms. The symbols used are taken from British Standard BS 2750: 1980. A more complete treatment can be found in CIRIA Report 114.

In buildings, sound insulation methods can be divided into two types:

Acoustic standards are a product of both physical needs i.e. the need to sleep or to hold a conversation, and the general expectancies of building users. It is generally acceptable, for instance, for a certain amount of sound to cross between a typical domestic kitchen and living room, however it is not acceptable in most western cultures for conversations to be audible between dwellings.

It is apparent that the standards of acoustic insulation that are required in different parts of a building will vary, and that the performance required of individual building elements will reflect this variation. For example, walls between offices and a workshop are likely to require to have greater acoustic insulation properties than those between a trade counter and the same workshop.

Airborne sound insulation

Airborne sound insulation between rooms can he measured by generating a steady sound of a particular frequency content in one room (the source room) and comparing it with sound in a second adjacent room (the receiving room). These measurements are made at a number of different frequencies.

The difference between the two levels is referred to as the level difference D. This level difference is influenced by the amount of acoustic absorption in the receiving room. The absorption can he estimated by measuring the reverberation time T - the time taken for the reverberant noise to decay by 60 dB. In order that measurements in different buildings may be compared, the level differences can be adjusted to a standard reverberation time of 0.5 s. This gives the standardised level difference DnT .

Individual building elements such as partitions, doors or windows can he tested in acoustic laboratories. These laboratories comprise two massively constructed adjacent rooms which are isolated against flanking transmission (see Section 2.2) and connected by an aperture containing a test panel of the building element. The level difference is measured between the two rooms and the result adjusted to be independent of both the area of the panel and the acoustic absorption of the room. The resulting value is the sound reduction index R.

Impact sound insulation

Impact insulation tends only to he relevant to floors. A standard impact sound source (tapping machine) is used, comprising a row of hammers which strike the floor repeatedly at a standard rate. The resulting sound in the receiving (downstairs) room is measured and this value is termed the impact sound pressure level L. Measurements in buildings can be standardised to a reverberation time of 0.5 s. This gives the standardised impact sound pressure level L'nT. Tests in laboratories, normalised for area and absorption give the normalised impact sound pressure level Ln. L'nT is therefore generally a field measurement, whilst Ln is a more absolute laboratory type measurement.

NOTE: this test method means that the better the impact sound insulation, the lower the value of L, L'nT or Ln.

Single figure rating values

Sound insulation is normally measured at a number of different frequencies - usually 16 one-third octave hands from 100 Hz to 3150 Hz (refer to the graph in Figure 3). However, for many purposes, including the requirements for dwellings given in the Building Regulations, a single figure rating is required. There are several methods of reducing the sound insulation values at the sixteen individual frequencies to a single figure value. An obvious method is to take the arithmetic mean, but very high levels of sound insulation at some frequencies can offset poor performance at others. The most common method of overcoming this is to compare the measured results with a set of sixteen reference results i.e. a reference curve. The rating is made by considering only those sound insulation values which fall short of the reference curve. In this way, one or two very good results have much less effect on the single figure value. The method used for airborne sound is given in Figure 3. A similar method is used for impact sound, and full details can he found in BS 5821: 1984.

A system using the same basic principles but a slightly different frequency range is used in the United States for generating values of Sound Transmission Class STC.

The resulting single figure value is called the standardised weighted level difference DnT,w when generated from DnT values; the weighted sound reduction Rw when generated from values of R; the standardised weighted impact sound pressure level L'nT,w when generated from values of L'nT; and the normalised weighted impact sound pressure level Ln,w when generated from values of Ln.

Figure 3 Calculation of the standardised weighted level difference DnT,w

Notes on the figure:

1) Measure DnT values at the 16 frequencies and assess these against values of the reference curve in an arbitrary position (Position 1 above) as given in BS 5821.

2) Identify those of the 16 frequencies where the measured DnT values are less than those of the reference curve and subtract these values from those of the reference curve to determine the adverse deviations in dB.

3) Sum all of the adverse deviations, i.e. sum of adverse deviations = 48 at Position 1 above.

4) Move the reference curve in 1 dB steps until the sum of the adverse deviations is as large as possible but not greater than 32 dB, i.e. sum of adverse deviations = 27 at Position 2 above.

(Note: the reference curve may have to moved either up or down in order to achieve the required result.)

5) The DnT value of the reference curve (in Position 2) at 500 Hz is the standardised weighted level difference DnT,w. If the adverse deviation at any frequency exceeds 8 dB this must be reported with the DnT,w.

For further details refer to BS 5821: 1984: Methods of rating the sound insulation in buildings and of building elements.

2.2 Factors affecting sound insulation

This section deals with those design details which control the sound insulation in buildings, particularly framed constructions. The design details for airborne and impact sound are similar but not identical.

When a room is separated from another room, airborne sound can travel by two routes: directly through the separating structure - direct transmission, and around the separating structure through adjacent building elements - flanking transmission. These routes are indicated in

Figure 4.

Sound insulation for both routes is controlled by the following three characteristics:

Direct transmission depends upon the properties of the separating wall or floor and can be predicted from laboratory measurements. Flanking transmission is more difficult to predict since it is influenced by the way in which the building elements are configured and detailed. It is notable that, in certain circumstances, such as where separating walls have a high standard of acoustic insulation but side walls are constructed to lower standards and are continuous between rooms, flanking transmission can account for the passage of more sound than direct transmission.

Sound transmission across a solid wall or a single skin partition will obey what is known as the mass law. This law may be expressed in a variety of ways and equations are presented in Appendices A.2.3 and A.2.2. In principle the law suggests that the sound insulation of a solid element will increase by 5 dB per doubling of mass.

The mass law does not however apply to lightweight framed constructions which achieve far better standards of sound insulation than the law would suggest owing to the presence of a cavity, and the degree of isolation that is achieved between the various layers of the construction. It has been demonstrated that the sound insulations of individual elements within a double skin partition tend to combine together in a simple cumulative linear relationship. The overall performance of a double skin partition can therefore generally be determined by simply adding together the sound insulation ratings of its constituent elements. In this way, two comparatively lightweight partitions of 25 to 30 dB sound reduction can be combined to give an acoustically enhanced partition with a 50 to 60 dB sound reduction, whereas the mass law alone would have suggested only a 5 dB improvement. This is the basis of many lightweight partition systems, and is further illustrated in Figure 5.

Figure 5

Particularly high levels of sound insulation can be achieved using two independent frames with an airspace between, each supporting one surface of the wall. Similarly, lightweight floors can be designed to incorporate an insulating layer within the construction of the deck which both absorbs a certain amount of sound and provides a high degree of isolation between the floor layers. Heavy concrete floors have a natural resistance to the transmission of sound, particularly when used with resilient layers (such as proprietary closed-cell foams).

The sound insulation properties of separating walls in framed buildings are very good, and often significantly better than equivalent masonry constructions.

It is important to provide adequate sealing around floors and partitions since even a small gap can lead to a marked deterioration in acoustic performance. Usually walls are sealed by the plaster finish; however, where walls abut profiled metal decks, or similar elements, sealants may be required. Where there are movement joints at the edges of walls, special details are likely to be necessary and advice should be sought from manufacturers.

Ideally, wall linings should not be penetrated by services. This is particularly important for separating walls between dwellings. Where service penetrations do occur in sensitive locations particular attention should be given to the way in which these are sealed.

Flanking transmission is difficult to predict. Even in the apparently simple case of homogeneous constructions of cast in situ unlined concrete, analysis is far from straightforward; however, certain facts have been established from experience.

The worst situation is where a solid lightweight masonry wall runs continuously past the end of a separating element. In this situation flanking transmission can be reduced by:

Introducing a physical break in the flanking wall at the junction with the separating element. Constraining movement in the flanking wall by tying it into the separating element. This can normally only be achieved where the separating element is of similar mass to the flanking wall. It is unrealistic for instance to tie a masonry flanking wall into a separating wall of lightweight framed construction. Fixing independent linings to the flanking walls similar to those which may be used to increase direct sound insulation.

Framed constructions are normally built of separate panels such that there is a natural junction between the flanking and separating elements. This junction, together with the fact that all wall and floor elements tend to be of similar mass, means that flanking transmission is much less of a problem than in many masonry constructions, and it is likely to be direct transmission which determines the requirements for sound insulation.

It has been found that in simple solid construction, flanking transmission depends on the length of the junction between the flanking wall and the separating wall, rather than the area of the flanking wall. An aperture in the flanking wall, such as a window, close to the junction with the separating wall, can reduce the effective length of this junction and thereby the flanking transmission.

Approved Document E to the Building Regulations for England and Wales details a number of forms of construction which meet the requirements of the Regulations for sound insulation. The descriptions of these constructions include limitations which apply to methods for reducing flanking transmission. Where framed walls or floors are supported on framed external walls, the Document requires very few measures in order to reduce flanking transmission.

Occasionally, resonances in panels can reduce sound insulation. All panel based systems will have some frequencies at which resonance occurs, and whilst this is not normally considered a major issue, it can be prudent to check that resonant frequencies are either well above or below those which are most audible to the human ear (approximately 100 Hz to 3150 Hz). Information on how to estimate resonant frequencies is given in Appendix A.2.4.

Additional factors for impact sound

Impact sound arises from a variety of sources, most notably the movement of people within a building, but also from such things as the slamming of doors and the use of electrical sockets. The only type of impact sound covered by regulatory control in most countries, including the UK, is that from footsteps. Specific precautions to avoid impact sound are normally therefore only associated with floor surfaces.

Impact sound can be reduced by the use of a very soft floor covering but the most common solution is to use a resilient layer under a hard floor surface. In framed constructions the floor surface is usually timber or chipboard, sometimes with a layer of plasterboard beneath to increase the mass. Beneath this surface is a resilient layer of mineral wool or expanded plastic foam which usually rests on the structural floor surface, although sometimes can rest directly on the top of the joists.

Specifying the appropriate resilient layer with the correct dynamic stiffness under the imposed loadings. Ensuring that the systems have sufficient durability (i.e. that resilient layers will not permanently compress during service, and that all components are sufficiently robust to cope with the anticipated floor loadings). Isolating the floating floor surface from the rest of the structure at the floor edges. This is normally achieved by returning the resilient layer up the edges of the walking surface.

Where comparatively lightweight floor constructions are used, as in most types of framed buildings, the mass of the floating layer can also have a significant effect on the airborne sound insulation.

2.3 Comparison of sound insulation of different construction methods

This section discusses the acoustic performance of different types of construction.

Construction quality has a major impact on the acoustic performance of building elements. Prototype elements may perform well under test conditions, but less well in completed buildings because of poor installation or construction detailing. The standards of acoustic insulation provided by a particular form of construction can however be determined by carrying out a large number of field tests such as those summarized in BRE Digest 333 Sound insulation of separating walls and floors, Part l: Walls.

Table 1 is reproduced from BRE Digest 333. It gives results for masonry and equivalent framed constructions using timber. From the table (Ref. No. 19) it can he seen that 99.9% of timber framed dwellings have a sound insulation of 52 DnT,w or better and therefore satisfy Approved Document E to the Building Regulations. Plastered solid brick walls have historically been used as the standard for sound insulation between attached dwellings. It can be seen that this wall type (Ref. No. 4) achieved a pass rate of 83%, - substantially less than framed dwellings. Similar findings are contained in two further BRE reports: Sewell 1977 and Sewell 1978

There is a considerable body of evidence to show that the sound insulation of steel framed dwellings is at least as good as that of equivalent timber framed dwellings (BRE Digest 347, ASTM E497 - 76 and Fontan, amongst others references, support this finding).

Direct comparisons of timber and steel framed panels have been made by British Gypsum Limited who have their own in-house sound insulation test facilities. Data available from British Gypsum has been reproduced in the form of two graphs in Figures 6 and 7. (Full details of the steel stud system and other steel framed partition systems are given in the British Gypsum 'White Book'.)

Figures 6 and 7 demonstrate that at most frequencies steel stud framing gives superior results to that of equivalent timber construction. This is in part attributable to the greater flexibility of steel studs, which serves to improve the isolation of elements within the partition.

Contents