Fire Design Concepts
3.5. Design model for the verification of the separating function
In order to limit fire spread by guaranteeing adequate fire compartmentalization, elements form-ing the boundaries of fire compartments are designed and constructed to maintain their separatform-ing function throughout the anticipated fire exposure, which places requirements on their integrity
Key:
1. Initial surface of member 2. Border of residual cross-section 3. Border of effective cross-section
d char,n notional charring depth
d 0 zero strength layer: d 0 = 7mm
Fig. 3.4: Definition of residual cross section and effective cross section [40]
3.5. DESIGN MODEL FOR THE VERIFICATION OF THE SEPARATING FUNCTION 31 ( E ) and insulation ( I ). The required period of time is normally expressed in terms of fire resis-tance using the standard fire exposure and is specified in building regulations. While fire tests are still used extensively for the verification of the separating function of timber substructures like light-frame wall-and-floor assemblies, design models are becoming increasingly common.
For the ISO-fire exposure criterion, I (insulation requirement) may be assumed to be satisfied, if the average temperature rise over the whole of the non-exposed surface is limited to 140°C, and the maximum temperature rise at any point of that surface does not exceed 180°C. This prevents ignition of objects in neighbouring compartments. The criterion E (integrity requirement) may be assumed to be satisfied, if no flames or hot gases on the non-exposed side of an assembly can be observed. As I is clearly defined, verification can be made by heat transfer calculations instead of testing. The E, on the other hand, is mostly determined by observations, because calculations are too complex and involve prediction of when cracks will form, hot gas dynam-ics, and other factors commonly beyond analytical capabilities. Premature integrity failure may occur because of sudden failure of claddings or opening of gaps, which often is dependent on how the material layers are fastened together. Extensive experience with full-scale testing of wall-and-floor assemblies supports rules about the construction detail ing of some wall-and-floor assemblies included in EN 1995-1-2, for example. The integrity criterion may be assumed to be satisfied, if the insulation criterion has been satisfied and panel materials remain fastened to timber frameworks on non-exposed sides of assemblies.
In timber buildings, walls and floors are mostly built up by adding different layers to form an assembly. For the verification of the separating function of timber assemblies, additive methods are common for determining combined behaviour of components. With additive models, the fire resistance of layered constructions is obtained by simply summing the contribution of the individual layers. Calculation models for verifying the contributions of the different layers to the separating function of light timber frame wall-and-floor assemblies are used in the UK [52], Canada [17] and Sweden [53]. The current design method according to EN 1995-1-2 (Annex E) is based on the Swedish component additive method. As an enhancement of the Canadian method, the Swedish approach considers the influence of adjacent materials on the fire perfor-mance of each layer and, therefore, more realistically describes the fire perforperfor-mance. However, it should be cautioned that the method is based on data from only a limited number of fire tests on wall assemblies and, therefore, only covers a limited range of timber structures. König et al.
[54] give a comprehensive review of the related international practices.
A comprehensive research project on the separating function of light timber frame wall-and-floor assemblies with cladding made of gypsum plasterboards and wood-based panels was car-ried out at ETH Zurich, in collaboration with Empa Dubendorf [55,56]. The objective was the development of an improved design model for the verification of the separating function ( E and I criteria) of light timber frame wall-and-floor assemblies. A large number of small-scale fire tests permitted the analysis of the influences of material type, thickness, position, and number of the layers on the thermal behaviour of protective cladding made of gypsum plasterboards and wood-based panels. Results of the fire tests allowed the verification and calibration of ther-mal properties used in therther-mal finite element (FE) analysis. An extensive FE parametric study enabled calibration of coefficients of a model used for the verification of the separating function of light timber frame wall-and-floor assemblies. The resulting model is capable of considering timber assemblies with an unlimited number of layers of gypsum plasterboards, wood panels, or combinations thereof, with wall cavities that may be either empty or filled with rock or glass
fibre insulation. The model is based on the additive component method given in EN 1995-1-2.
Thus, the fire resistance t ins of the timber assembly is taken as the sum of the contributions from the different layers for the worst possible path of heat transfer (Fig. 3.5), and according to their function and interaction (Fig. 3.6 ):
= = Sum of the protection values t prot ,i of the layers (in the direction of the heat flux) preceding the last layer of the assembly on the fire-unexposed side [minute]
t ins,n = Insulation value t ins,n of the last layer of the assembly on the non-exposed side [minute]
Protection and insulation values of the layers can be calculated according to Eqs. (3.2) and (3.3), taken into account the values of the layers:
t prot ,i = (t prot ,0,i · k pos,exp,i · k pos,unexp,i +∆t i) · k j,i (3.2)
t ins,n = (t ins,0,n · k pos,exp,n, + ∆t n) · k j,n (3.3)
with t prot,0,i= basic protection value [minute] of layer i, see Fig. 3.6
a e a–e Heat transfer paths
Fig. 3.5: Possible heat transfer paths through separating multiple layer construction
Timber frame member
Fig. 3.6: Timber frame wall-and-floor assemblies: numbering and function of the different layers
3.5. DESIGN MODEL FOR THE VERIFICATION OF THE SEPARATING FUNCTION 33
t ins,0,n= basic insulation value [minute] of the last layer n of the assembly on the
non-unex-posed side, see Fig. 3.6
∆t i, ∆t n = correction time [min] for layers protected by gypsum plasterboards of type F or type× as well as gypsum fibreboards (GF)
k pos,exp,i, k pos,exp,n = position coefficient that takes into account the influence of layers preced-ing the layer considered
k pos,unexp,i = position coefficient that takes into account the influence of layers backing the
layer considered k j,i, k j,n = joint coefficient
The basic insulation value t ins,0 corresponds to the fire resistance of a single layer without the influence of adjacent materials and joints (i.e. the average temperature rise over the whole of the non-exposed surface is limited to 140°C, and the maximum temperature rise at any point of that surface does not exceed 180°C). The basic insulation value can be assessed by tests or FE thermal analysis. For FE thermal analysis, only the temperature criterion of 140°C is used.
The temperature of the layer at the beginning of the analysis on the exposed side and the non-exposed side is assumed to be 20°C. The definition of the basic insulation value t ins,0 is illustrated in Fig. 3.7.
Wall-and-floor assemblies with only one layer are rarely used in buildings. Most assemblies consist of two or more layers. The contribution to the separating function of the construction of each layer, except the last layer of the assembly on the exposed side, is mainly protection of the layer(s) below (see Fig. 3.6 ). Therefore, it is appropriate to introduce a basic protection value
t prot,0 defined as the time until failure of the protective function. Analogous to the calculat ion for
fire protective claddings of load-bearing timber constructions, according to EN 13501-2 [57], the definition of the basic protection value t prot,0 is as illustrated in Fig. 3.8. The testing method for fire protective claddings is performed with a particleboard with a thickness of 19 mm back-ing the layer studied. The contribution of the claddback-ing to the fire protection of the particleboard may be assumed to be satisfied, if the average temperature rise over the whole exposed surface of the particleboard is limited to 250°C and the maximum temperature rise at any point on that surface does not exceed 270°C. For FE thermal analysis, only the temperature criterion of 250°C is used. The temperature of the layers on the exposed and non-exposed sides at the begin-ning of an analysis is assumed to be 20°C (Fig. 3.8). Analytical equations for the calculation of the basic insulation value t ins,0 and the basic protection value t prot,0 for different materials have been calculated by FE simulations and verified by fire tests.
The position coefficient considers the position of a layer within an assembly, in the direction of the heat flux, because the layers preceding and backing a layer under consideration have an influence on its fire behaviour. The physical meaning of the position coefficient is illustrated in
Considered layer
Fig. 3.7: Definition of the basic insulation value t ins,0 using FE thermal analysis
Fig. 3.9 for a three-layer assembly. For simplicity, it is assumed that each layer has the same thickness and density and that the influence of joints between layers is neglected. In this case, the basic protection value for each layer is the same (t prot,0,1 = t prot,0,2 = t prot,0,3). The first layer is directly exposed to fire and backed by the second layer. The temperature of all layers at the beginning of the fire on the exposed side and the non-exposed side is assumed to be 20°C (Fig. 3.9a). The contribution of the first layer to the total fire resistance is defined as t prot,1. The position coefficient k pos,1 of the first layer can be described as the ratio t prot,1 to t prot,0,1 and depends on the layer backing the first layer. The second layer is protected by the first layer. It is conservatively assumed that, after failure of the protection provided by the first layer that the second layer is directly exposed to fire, with failure occurring when temperature at the interface between the first and second layers reaches 270°C time t = t prot,1. The main differences in com-parison with the initially unprotected first layer is that the temperature in the fire compartment is already at a high level and that no protective char layer exists [45]. Further, the temperature of the second layer on the fire-exposed side is 270°C, as defined previously, while the temperature on the non-exposed side is equal or greater than 20°C depending on the thickness of the second layer and the material preceding and backing the layer (Fig. 3.9b). For these reasons, the con-tribution of the second layer to the total fire resistance is lower than the concon-tribution of the first layer, that is, t prot,2 < t prot,1. The position coefficient k pos,2 of the second layer can be described as the ratio t prot,2 to t prot,0,2and is < 1.0. For the same reasons, EN 1995-1-2 assumes that after failure of a cladding, charring occurs at the increased rate of initially unprotected surfaces [41]. The third layer is protected by the second layer. After failure of the protection by the second layer at 270°C time t = t prot,1 + t prot,2, the third layer is directly exposed to fire.
20°C + 250°C = 270°C
Fig. 3.8: Definition of the basic protection value t prot,0 using FE thermal analysis
h
(b) Second layer exposed to fire:
t = t prot,1
(c) Third layer exposed to fire:
t = t prot,1+ t prot,2
Fig. 3.9: Temperature distribution at different times of timber assembly with three layers
3.5. DESIGN MODEL FOR THE VERIFICATION OF THE SEPARATING FUNCTION 35 The third layer is the last layer in the assembly and takes an insulating function (Fig. 3.6 ). There-fore, for that layer, the temperature criterion of 140°C is applicable, and an insulation value of
t ins,3 must be calculated. Because of the further increased temperature in the fire compartment
and the missing protective char layer, as well as the temperature rise criterion of 140°C, instead of 250°C, the contribution of the third layer to the total fire resistance is lower than the contri-butions of the first and second layers, that is, t ins,3 < t prot,2 < t prot,1. The position coefficient k pos,3 of the third layer can be described as the ratio t ins,3 to t ins,0,3 and is < 1.0. The fire resistance of the entire timber assembly is the sum of the contributions from the different layers, that is, t ins = t prot,1 + t prot,2 + t ins,3.
The influence of the layers preceding and backing the layer considered is analysed separately.
In that case, the position coefficient k pos,exp accounts for the influence of the layers preceding a layer considered, whereas the influence of the
layer backing the layer studied is considered by k pos,unexp. The position coefficients calculated using FE simulations showed that the position coefficient k pos,exp is mainly influenced by the fire temperature at the time when a layer being con-sidered is exposed directly to the fire, as opposed to the thickness of the layer itself. The influence of the preheating was shown to be small. The design model for the verification of the separat-ing function of timber assemblies was developed for the ISO fire exposure and is a function of the sum of the protection values of the preceding layers (i.e.
∑
t prot,i−1). The thickness of the layer considered is expressed as the basic protection value t prot,0,i or basic insulation value t ins,0,n. As an example, Fig. 3.10 shows the position coefficientk pos,exp of gypsum boards with different
thick-nesses, as a function of the sum of the protection values of the layers preceding the gypsum board (i.e.
∑
t prot,i−1). The position coefficient k pos,exp decreases with increasing∑
t prot,i−1, because thegypsum board is protected longer and at higher exposure temperatures. The position coefficient k pos,exp decreases with reducing thicknesses of the gypsum board. Similar effects are observed for wood-based panels and batt cavity insulation, making it possible to determine the position coefficient k pos,exp as a function of the sum of the protection values of the layers preceding a layer (
∑
t prot,i−1) and the basic values of a layer considered (t prot,0,i and t ins,0,n). Calculation of position coefficient k pos,exp is therefore simple.The influence of a layer backing the layer of interest is defined by the position coefficientk pos,unexp. Results of fire tests, Fig. 3.11, supported by FE simulations show that the influence of a backing layer is small, if the backing layer is made of gypsum or timber. Also, batt insulation backing layers cause other layers to heat up more rapidly reducing the protection time of the layer, and therefore the position coefficient k pos,unexp should be evaluated for the case of batt insulation back-ing. For timber or gypsum backing assuming k pos,unexp,i = 1.0 is conservative.
00
Protection timeΣtprot,i-1 (min)
P o s gypsum boards with different thickness as a function of the sum of the protection values of the layers preceding the gyp-sum board (i.e.
∑
t prot,i−1)The methods described in this chapter extend the range of application of EN 1995-1-2 approaches significantly, with further details given elsewhere [55,56].