- Email: [email protected]
Experimental investigation of the fire resistance of multi-layer drywall systems incorporating Vacuum Insulation Panels and Phase Change Materials
Fire Safety Journal 81 (2016) 8–16
Contents lists available at ScienceDirect
Fire Safety Journal journal homepage: www.elsevier.com/locate/firesaf
Experimental investigation of the fire resistance of multi-layer drywall systems incorporating Vacuum Insulation Panels and Phase Change Materials Dimos A. Kontogeorgos n, Georgios K. Semitelos, Ioannis D. Mandilaras, Maria A. Founti National Technical University of Athens, School of Mechanical Engineering, Laboratory of Heterogeneous Mixtures & Combustion Systems, Heroon Polytechniou 9, Zografou Campus, Athens 15780, Greece
art ic l e i nf o
a b s t r a c t
Article history: Received 10 September 2015 Received in revised form 22 January 2016 Accepted 27 January 2016
This paper studies the fire resistance of innovative high thermally insulated multilayer drywall assemblies incorporating conventional insulation materials, Phase Change Materials (PCMs) and Vacuum Insulation Panels (VIPs). An experimental study was developed and implemented into two directions. In the first direction, four different multilayer drywall configurations were subjected to fire temperatures up to 900 °C from one side, while the other side was at ambient conditions. Each configuration consisted of a gypsum board with PCMs (PCM-GB), a standard gypsum board (S-GB), an Expanded Polystyrene (EPS) layer, a thermal insulation render containing EPS (TIR) and an insulation layer located between the PCM-GB and the S-GB. A different insulation layer was used for each configuration: cavity (no insulation), EPS, mineral wool (MW) (both conventional insulation materials) and Vacuum Insulation Panels (VIP) (super insulation material). In the second direction, Differential Scanning Calorimetry (DSC) measurements, at inert (nitrogen) and oxidized (air) environments, were performed for all the utilized materials. DSC results indicated that at temperatures up to 200 °C, the gypsum boards (both PCM-GB and S-GB) act as fire retardants because of the dehydration process. The paraffin and PMMA components of the PCM started to evaporate and oxidize at temperatures higher than 200 °C and up to 500 °C. The resin binder of the mineral wool started to volatilize and oxidize at 265 °C, while at 500 °C the mineral wool started to melt. The volatilization of the EPS started at 275 °C, while the full volatilization and oxidation took place at the temperature range between 420 °C and 550 °C. The chemically bound water of the TIR dehydrated at the temperature range between 50 °C and 200 °C, while the EPS contained in the TIR behaved similar as the EPS sample. Finally, the cellulose fibers contained in the VIP volatilized and oxidized at the temperature range between 320 °C and 480 °C. Furnace results confirmed the fire resistance behavior of the gypsum boards indicated by the dehydration “plateau”. The wall assembly with the EPS layer found to behave similar to the assembly with the cavity due to the fact that the EPS melted at temperatures near 200 °C. The wall assembly with the mineral wool delayed the temperature rise until 500 °C where it started to melt. The VIP layer found to significantly delay the penetration of the heat through the drywall configuration when compared to the other configurations. According to the failure criteria regarding excessive temperature rise on the ambient facing side of the wall, the VIP layer was found to increase the time-to-failure by approximately 68%, with respect to the assembly with the cavity. On the other hand, the respective time increase for the conventional insulation materials was 2% and 19% for the EPS and the MW, respectively. & 2016 Elsevier Ltd. All rights reserved.
Keywords: Fire resistance Fire behavior VIP PCM EPS Mineral wool Super insulation materials DSC
1. Introduction In the last two decades, the need for sustainable buildings that minimize the consumption of raw materials and energy [1] has n Correspondence to: Laboratory of Heterogeneous Mixtures and Combustion Systems, Thermal Engineering Section, School of Mechanical Engineering, National Technical University of Athens, Heroon Polytechniou 9, Polytechnioupoli-Zografou, Athens 15780, Greece. E-mail address: [email protected] (D.A. Kontogeorgos).
http://dx.doi.org/10.1016/j.firesaf.2016.01.012 0379-7112/& 2016 Elsevier Ltd. All rights reserved.
gradually shifted the focus of the building sector towards lightweight constructions and nearly Zero Energy Buildings (nZEB). Lightweight multi-level buildings based on Dry Wall Systems (DWS) provide a safe, fast and long lasting solution to housing, particularly in high risk areas, such as highly seismic areas [2,3]. On the other hand, the presence of low density materials and thin envelopes in the lightweight buildings reduces the thermal efficiency (low thermal mass and low insulation) of the buildings. The latter, creates a number of new challenges for the design of
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
Nomenclature Abbreviations DSC DWS EPS MW nZEB
Differential Scanning Calorimetry Dry Wall System Expanded Polystyrene Mineral Wool Nearly Zero Energy Building
lightweight buildings. An attractive solution to this challenge appears to be the efficient integration of Phase Change Materials (PCMs) and Super Insulation Materials (SIM) in the lightweight building envelope [4]. The PCMs, due to their high amount of latent heat, utilized for thermal energy storage in building elements, are able to cope with thermal and electricity peak loads [5]. The SIMs, due to their very low thermal conductivity, provide efficient wall insulation solutions, as well as efficient solutions to thermal bridging effects [6]. Hence, the incorporation of PCMs and SIMs in DWS can lead to sustainable and thermally efficient lightweight constructions. Thermal Energy Storage (TES) in general and PCMs in particular, have been a research topic for the last 20 years. TES is an area of international interest dealing with energy saving, efficient and rational use of available resources [7–12]. It provides solutions in very specific areas: time delay and available power between production or availability of energy and its consumption in receiving systems (solar energy, cogeneration, etc.); security of energy supply (hospitals, computer centers, etc.); thermal inertia and thermal protection of the buildings. The thermal inertia and thermal protection are the areas where the PCMs have achieved a relatively high market penetration. Applications include, among others, their incorporation in the core of building materials. Particularly in the building sector, PCMs have been successfully integrated in the building fabric either by directly incorporating PCM microcapsules into commonly used building materials, such as in concrete, gypsum boards and natural stones [13–15] or by using separate layers of shape stabilized PCMs in wall assemblies [16]. The utilization of PCMs could lead to reduction of power requirements of the heating and cooling equipment commonly used to secure continuous energy supply. In the building context, the incorporation of PCMs into the core of building materials is intended to increase the thermal storage capacity of the building element [17]. The technology takes advantage of the latent heat of the PCM during the solid–liquid change of state to stabilize the temperature of the material and reduce the heat losses/gains from the building to the environment [18,19]. However, PCMs are usually mixtures of hydrocarbon molecules, which at temperatures higher than 200 °C evaporate. The gas mixture produced is flammable and may affect the fire endurance of the building element [20]. SIMs and particularly the Vacuum Insulation Panels (VIP) have been developed in the last decade, mainly for use in appliances, such as refrigerators and deep-freezers [21,22]. The reason hereto is the significantly low thermal conductivity (5–7 mW/(m K)), a factor of five to eight times better than the conventional insulation materials (Z 30 mW/(m K)) [23]. In the frame of building constructions, VIPs enable thin and highly insulation constructions to be realized for walls, floors and roofs. The motivation for examining the SIMs in buildings comes mainly from the difficulties involved in renovation, aesthetic considerations, as well as from the necessity to reduce the overall energy consumption of the buildings. The latter is accomplished by increasing the thermal resistance of the envelope's walls, as well as by reducing the effect
PCM PCM-GB PFP S-GB SIM TES TIR VIP
9
Phase Change Materials Gypsum board with Phase Change Materials Passive Fire Protection Standard gypsum board Super Insulation Material Thermal Energy Storage Thermal Insulation Render Vacuum Insulation Panel
of thermal bridges introduced when materials with higher thermal conductivity (e.g. steel studs) break through or partially break through a layer with higher thermal resistance (i.e. lower thermal conductivity) creating pathways of lower thermal resistance causing increased heat losses [24–26]. VIPs consist of inorganic compounds and thus, it is expected that they would enhance the fire resistance of a wall configuration. To the authors' knowledge, there is not any research found in the literature on the fire behavior of VIPs. Fire protection in the building technology is still extremely topical, especially due to the increase of the legislation and fire standard restrictions [27,28]. Dry wall systems (e.g. gypsum boards, cement boards etc.) are the most common systems used in lightweight steel skeleton buildings as Passive Fire Protection (PFP) systems for the steel structure (e.g. frames, studs etc.) [29]. This is due to the fact that they have very good fire behavior, associated with the water contained in their crystal structure, which is evaporated under fire conditions absorbing significant heat quantities from the fire and thus, delaying the heat penetration through the assembly [30–32]. The objective of the current work is to study for the first time the fire resistance of innovative and high thermally insulated wall assemblies in terms of energy storage and super insulation capabilities. The examined combinations of multilayer assemblies can find direct application in drywall construction due to their exceptionally good thermal insulation properties and relatively low thickness. An experimental study was developed and implemented, where “small-scale” furnace tests, as well as DSC measurements were performed in order to deeply analyze the fire resistance of the examined configurations. Within the furnace tests, four different multi-layer drywall configurations incorporating PCMs, VIPs and conventional insulation materials (mineral wool and expanded polystyrene) were subjected to fire temperatures up to 900 °C from one side, while the other side was at ambient conditions. Additional DSC measurements for all the components of the examined assemblies were performed in order to assess their thermal degradation at temperatures up to 600 °C. The main innovation of this work is owed to the study of the fire behavior of VIPs and PCMs, the interaction between different building materials that form a wall assembly under fire conditions and to the provision of experimental data in order to calibrate and validate dedicated fire numerical models.
2. Experimental study The overall experimental strategy followed two directions. Firstly, the chemical reactivity of the examined materials under inert and oxidized conditions was examined by performing DSC measurements. Secondly, “small-scale” furnace experiments were performed, in which multi-layer configurations, composed of different building materials, were positioned in front of a high temperature radiation furnace. The temperature evolution at different positions through the configurations was recorded, and together
10
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
Table 1 General information about the utilized materials. Material
k (W m
S-GB PCM-GB Mineral Wool EPS VIP
0.25 0.28 0.035 0.032 0.007
1
K
1
)
ρ (kg m
3
)
820 790 35 40 195
Cp (J kg
1
K
1
)
Composition – 19.1% encapsulated PCM (78% n-heptadecane with melting point near 22 °C, 22% PMMA) – – 98% fumed silica 2% cellulose fibers, organic binders and graphite based opacifiers
1000 1000 850 1300 800
subtracted from the DSC curve measured by having placed a sample mass in the same pan. In order to define the baseline at the temperature range where each reaction takes place, integral tangential baselines, provided from the apparatus software, were utilized. Finally, the average mass of the examined samples used in the DSC measurements was near 20 mg for the gypsum boards and the thermal insulation render and around 8 mg for the insulation materials (mineral wool, EPS and VIP). Each DSC measurement was repeated three times and the reproducibility error found to be within the instrument limits (i.e. 72%). 2.3. Furnace experiments
Fig. 1. Schematic diagram of the multi-layer wall configurations.
with the DSC results, their fire response and fire resistance was assessed. 2.1. Materials The materials that were used for all the experiments (DSC and “small-scale” furnace experiments) were commercially available. Table 1 summarizes the main general information of the utilized materials. 2.2. DSC experiments The chemical reactivity of the examined samples was defined by DSC measurements in a Stare SW 8.10, Mettler Toledo apparatus with an accuracy of 72%. The measurements were performed at the temperature range between 25 °C and 600 °C, using 10 °C/min and 20 °C/min heating rates, in inert (nitrogen) and oxidized (air) atmospheres, with a gas flow of 100 ml/min, using 20 μl aluminum crucibles. The apparatus was firstly calibrated at 156.6 °C using indium and at 419.6 °C using zinc, while before each measurement a blank DSC run was performed with an empty pan using the same conditions. The resulting blank curve was
The second part of the experimental study comprised four different “small-scale” furnace experiments. The experiments were designed in order to study the behavior of different building materials used in DWS and their interaction when exposed to fire conditions, as well as to serve as benchmark regarding the verification of detailed numerical fire models. More specifically, four different multilayer wall configurations, each consisting of five layers, were subjected to fire conditions from one side (i.e. furnace entrance), while the other side was at ambient conditions. According to standard ASTM E 119-00a [33] there is a degree of repeatability and reproducibility of fire testing. However, results depend on several factors, such as the type of assembly and material being tested, the nature of the boundary conditions and the details of the workmanship during the preparation of the specimens. For that reasons, it is impossible to quantify the anticipated error and thus, each of the “small-scale” furnace experiments was performed once. 2.3.1. Test specimens Fig. 1 illustrates a schematic diagram of the multi-layer wall configurations. As it is shown, beginning from the left side of the wall assembly, which is the layer that faces the furnace's entrance, the first layer is a gypsum board with PCMs (PCM-GB), the third layer is a standard gypsum board (S-GB), the fourth layer is an EPS and the fifth layer is a thermal insulation render layer containing EPS (TIR). The only difference between the experiments was that the second layer was different for each experiment; there was a cavity (no insulation) in experiment 1, an EPS in experiment 2, a Mineral Wool (MW) in experiment 3 and a VIP in experiment 4. The total thickness of all the configurations examined was 50 mm. Table 2 summarizes the list of the “small-scale” furnace experiments, as well as the materials used in each configuration and their thicknesses.
Table 2 List of performed experiments and utilized materials. Exp. number
Layer 1
Experiment Experiment Experiment Experiment
PCM-GB PCM-GB PCM-GB PCM-GB
1 2 3 4
(12.5 mm) (12.5 mm) (12.5 mm) (12.5 mm)
Layer 2
Layer 3
Layer 4
Layer 5
Cavity (12.5 mm) EPS (12.0 mm) MW (12.5 mm) VIP (10.0 mm)
S-GB S-GB S-GB S-GB
EPS EPS EPS EPS
TIR TIR TIR TIR
(12.5 mm) (12.5 mm) (12.5 mm) (12.5 mm)
(6.0 mm) (6.0 mm) (6.0 mm) (6.0 mm)
(6.5 mm) (7.0 mm) (6.5 mm) (6.5 mm)
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
Fig. 2. Specimen holder for the “small-scale” furnace experiments including the special ceramic insulation and the metal frame.
2.3.2. Specimen preparation As mentioned above, the only difference between each “smallscale” experiment was that the second layer, beginning from the layer exposed to the furnace entrance, was different for each experiment, while all the other layers were the same. The main idea of the “small-scale” experiments was the placement of each specimen in front of the entrance of a furnace. In order to fix each specimen in front of the furnace, it was positioned within a special specimen holder, which was consisted of a special ceramic insulation for high temperatures held by a metal frame, as shown in Fig. 2. The ceramic insulation was used in order to ensure the one
11
dimensionality of the heat transfer and minimize the lateral thermal losses. Fig. 3 shows how the materials, equipped with thermocouple sensors, were positioned inside the ceramic insulation The small gaps that occurred between the ceramic insulation and each material were filled in by means of a special ceramic paste that withstands temperatures up to 1300 °C. Each specimen was equipped with twenty K-type thermocouples with an accuracy of 70.4%, allowing the measurement of local temperatures at different locations within the specimen and on its surfaces, as well as of the surrounding air inside and outside the furnace. The thermocouples with the minimum possible diameter that withstand the temperature field at each position were chosen. The thermocouple sensors used for measuring the temperature evolution inside the furnace, on the surfaces of the PCM-GB and on the hot surface of the S-GB were thick (0.24 mm diameter) in order to withstand the high temperatures. On the other hand, the thermocouple sensors used for measuring the temperature evolution at the ambient and on the remaining surfaces were fine (0.12 mm diameter). The measuring tips of the thermocouple sensors within the wall assembly were placed in grooves less than 0.5 mm deep at the surfaces of the gypsum boards (both PCM-GB and S-GB), the EPS and the TIR. The good thermal contact of the tips with the materials during the measuring period was assured with the use of a special fire resistant adhesive (see Fig. 3). The thin adhesive layer, which covered the tips of the thermocouple sensors, was prevented the exposure of the tips, located on the exposed surface, from the furnace's radiation. Thus, no temperature correction for radiation losses was needed. Moreover, due to the fact that the radiation properties of the adhesive were similar to the respective radiation properties of the materials, the thermal radiation load on the exposed and unexposed surfaces of the wall configuration was
Fig. 3. Materials equipped with thermocouples positioned inside the ceramic insulation.
12
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
Fig. 4. Schematic diagram for the location of the thermocouples for each interface of the multi-layer wall configuration.
not affected by the thin adhesive layer. Fig. 4 visualizes the location of the thermocouples for each interface between the materials, which composed the specimen. For the temperature measurements, an Agilent data acquisition unit with an accuracy of 72% was utilized, while the temperatures were recorded and stored in intervals of 2s using the ‘LabVIEW’ software. 2.3.3. Experimental methodology Each specimen (i.e. multilayer wall configuration and specimen holder) was positioned in front of a Nabertherms B180 radiation furnace (250 mm height, 320 mm width and 550 mm depth), in such a way so that one surface of the specimen (PCM-GB) to be exposed to the furnace's thermal load and the other to be at ambient conditions. After the placement of the specimen in front of the furnace, the furnace was turned on, and the temperature value of 900 °C was set as the target temperature. The experiment was terminated when the temperature values recorded by the thermocouples were constant, reaching steady state conditions. Fig. 5 illustrates the overall experimental set-up and the specimen positioned in front of the furnace.
3. Observations, results and discussion 3.1. DSC experiments Fig. 6 illustrates the DSC results for all the examined materials at inert (N2) and oxidized (Air) environments. The respective
Fig. 5. Experimental set-up and specimen (including the specimen holder) positioned in front of the furnace.
results for the PCM-GB samples are shown in Fig. 6a. The typical GB dehydration characteristic peaks (1A and 1B) are observed at the temperature range between 100 °C and 200 °C, regardless the utilized atmosphere [34,35]. This process is endothermic, absorbs significant amounts of energy and it is thus, capable of slowing down the fire spread through the examined materials. Phenomena associated with the PCM compounds are also observed. The DSC
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
13
Fig. 6. DSC measurements of the core material at inert and oxidizes environments: (a) gypsum board with PCMs (PCM-GB), (b) standard gypsum board (S-GB), (c) mineral wool (MW), (d) expanded polystyrene (EPS), (e) thermal insulation render (TIR) and (f) Vacuum Insulation Panel (VIP).
analysis at inert and oxidized environment revealed the effect of PCM incorporation into the samples. At inert environment, the paraffin and PMMA, which compose the PCM, evaporate in the temperature range between 200 °C and 500 °C, as indicated by peaks 4A and 4B, while under oxidized environment the paraffin and PMMA oxidize in the same temperature range, as indicated by peaks 3A and 3B. Fig. 6b illustrates the DSC results of the S-GB samples. Once again, the typical dehydration process (peaks 1A and 1B) is observed regardless the utilized atmosphere. At temperatures near 380 °C, a small exothermic peak is observed, which corresponds to the reorganization of the GB's crystal mesh (peak 2) [34,35]. The latter exists but not observed in Fig. 6a at the oxidized atmosphere due to the presence of the exothermic behaviour of the PCM's paraffin and the PMMA at this temperature range. Fig. 6c illustrates the DSC results for the mineral wool samples. At inert conditions, a slightly endothermic reaction is observed, as indicated by the peak shown at 265 °C. This reaction is associated to the volatilization of the resin binder inside the mineral wool. On the other hand, at oxidized conditions the volatiles of this resin binder are oxidized, as indicated by peak 5 in Fig. 6c. Finally, temperature 500 °C is the onset temperature where the material begins to melt [36]. Fig. 6d illustrates the DSC results for the EPS. The endothermic reaction observed at inert environment (peak 6A) is owed to the break of the chemical bonds inside the EPS [37]. When the samples are heated at temperatures around 160 °C, the sample begins to melt. Under inert conditions the evaporation of the polymer starts at around 275 °C, while at temperatures between 420 °C and 450 °C the polymer has almost completely evaporated (peak 6A). The released volatiles oxidize in the presence of air (peak 7A). It should be noted that the released energy due to the oxidation of
the EPS at the temperature range between 275 °C and 450 °C is less than the absorbed energy due to the evaporation of the EPS and thus, the oxidation process is still an “endothermic” process at this temperature range (peak 7A). At the temperature range between 450 °C and 550 °C the oxidation of the EPS is clearly an exothermic process, as shown from peak 7B. Fig. 6e illustrates the DSC results for the thermal insulation render. The examined render is a mixture of cement and EPS. Thus, at the temperature range between 50 °C and 200 °C the endothermic dehydration process of the chemically bound water of the cement is observed (peak 9) [38]. The peaks shown near 300 °C and 450 °C are associated to the evaporation of the EPS (peaks 6A and 6B) and to the oxidation of the EPS’s volatiles (peaks 8A and 7B). Finally, the DSC results for the VIP samples (i.e. core material) are shown in Fig. 6f. As it is displayed, an endothermic process is observed at the temperature range between 20 °C and 100 °C (peak 10) regardless the utilized atmosphere, which is related to the loss of water, which is contained into the cellulose fibers [39]. Moreover, at temperatures near 320 °C and 480 °C, respectively, two peaks are observed at both inert and oxidized atmospheres. Peaks 11A and 11B are associated to the volatilization of the cellulose fibers, which is an endothermic process, while peaks 12A and 12B are associated to the oxidation of the cellulose fibers volatiles [39]. 3.2. Furnace experiments Fig. 7 shows the temperature evolution on the exposed surface, at the interfaces and on the unexposed surface of the furnace experiments, as well as the temperature evolution inside the furnace and at the ambient, as measured by the installed
14
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
Fig. 7. Temperature evolution: (a) inside the furnace, (b) on the exposed side, (c) at the interface between the PCM-GB and the insulation layer, (d) at the interface between the insulation layer and the S-GB, (e) at the interface between the S-GB and the EPS, (f) at the interface between the EPS and the TIR, (g) on the unexposed side and (h) at the ambient.
thermocouples. As mentioned above, the measuring tip of the thermocouple used for measuring the temperature evolution inside the furnace was thick in order to withstand the high temperatures. Therefore, the response time was high and attenuated all possible fluctuation (if any) of the temperature. On the other hand, the thermocouple used for measuring the temperature evolution of the ambient air was thin with very fast response. Therefore, what it appears as noise in Fig. 7h is actually the fluctuation of temperature due to the movement of the air near the surface of the wall assembly caused by the buoyant forces. It is clearly shown that the dehydration process of both PCMGB and S-GB significantly delays the temperature increase through the assembly's thickness. The latter is indicated by the observed temperature “plateau” shown at the interfaces between the different layers (Fig. 7c–g), and is more evident in the unexposed side of the assembly (Fig. 7g), where the temperature rise reduces considerably for a major period of time (∼30 min for the cavity case up to ∼70 min for the VIP case). When the dehydration process ends, there is a temperature increase at the positions behind the S-GB and up to the unexposed side (Fig. 7e–g). This can be explained by the fact that during the dehydration process, the dehydrated layers, behind the dehydration front, have significantly higher temperature than the non-dehydrated layers. Thus, when the dehydration effect ends, the low temperature layers, protected by the dehydration front, suddenly face significantly increased temperatures. The observed peaks shown in Fig. 7e–g are associated to the oxidation of the EPS volatiles, which takes place in the region near the EPS and TIR layers in the temperature range between 300 °C and 550 °C, in line with the DSC results (Fig. 6d and e). Thus, during the specific experiments, performed under ambient
oxidized conditions, the EPS's volatiles oxidized increasing the temperature recorded in the thermocouples. This temperature increase affects the temperature evolution at the interfaces in front of the EPS layer, as clearly shown by the temperature peaks in Fig. 7b–d. Finally, there was no indication of the PCM oxidation, as showed in the DSC results under oxidized conditions, due to the under-ventilated conditions inside the radiation furnace. The specimen with the VIP in the second layer (experiment 4) has a significantly better fire performance than the other assemblies with conventional insulation materials or with the cavity. The utilization of a VIP layer leads to a major delay of the heat penetration through the materials located between the VIP and up to the unexposed side of the assembly. This is clearly shown in Fig. 7e–g where the temperature “plateau” of experiment 4 is much more pronounced than the respective “plateau” observed in the other experiments. On the other hand, the conventional insulation materials utilized in the second layer (i.e. mineral wool and EPS), are slightly protecting the wall assembly. Regarding the EPS insulation case (experiment 2), due to the fact that it melts at around 200 °C, the specimen behaves almost the same as the specimen with the cavity (experiment 1), as shown in Fig. 7c–g. This means that the EPS does not offer any fire protection to the wall assembly. Regarding the Mineral Wool (experiment 3), due to the fact that its components melt near 500 °C, the wall assembly is protected until this temperature, as depicted from Fig. 7d–g, where the observed temperature “plateau” is maintained for a longer period of time than the respective “plateau” of experiments 1 (cavity) and 2 (EPS). The above observations are also shown in Fig. 8, which illustrates the infrared images of the unexposed side for all the “smallscale” experiments at different time intervals. As it is shown, there
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
15
reaches approximately 68% (with respect to the case with a cavity but no insulation material), while the respective increase percentages for the conventional insulation materials are near 2% for the EPS and 19% for the Mineral Wool.
4. Conclusions
Fig. 8. Infrared images of the unexposed side of all the examined configurations at 30 min time intervals.
Table 3 Time-to-failure values for each experiment. Experiment Experiment Experiment Experiment Experiment
1 2 3 4
Time-to-failure (min)
Increase (%)
80.1 81.6 95.2 134.9
– 2 19 68
is a small error introduced (estimated ∼5%) in the central part of the assembly due to the lateral heat from the ceramic insulation of the special holder. The latter proves the one dimensionality of the performed experiments and certifies the applicability of the measurements for validating numerical fire models. Regarding the behavior of the materials, it can be seen that the specimen with the VIP (experiment 4) delays significantly the heat penetration through the assembly. The latter is more obvious from the infrared images and the temperature distribution on the unexposed surface at 90 min, 120 min, 150 min and 180 min. The specimen with the EPS (experiment 2) behaves almost the same as the specimen with the cavity (experiment 1), as shown by the temperature distribution on the unexposed surfaces. Finally, the specimen with the mineral wool (experiment 3) behaves better that the specimens with the cavity (experiment 1) and the EPS (experiment 2), but not better than the specimen with the VIP (experiment 4). The timeto-failure values were calculated for each experiment, according to the failure criteria of the excessive temperature rise on the ambient facing side of the wall [40], which is valid when the temperature rise on the ambient facing side is a maximum of 180 °C or an average rise of 140 °C. The calculated results are tabulated in Table 3. It is obvious that the utilization of a VIP instead of a conventional insulation material leads to a significant increase of the time-to-failure. In this case, the time-to-failure increase
In this paper, an experimental study was developed and implemented in order to investigate the fire resistance and behaviour of innovative, high thermally insulated, multilayer drywall assemblies incorporating conventional insulation materials, VIPs and PCMs when exposed to fire conditions. Four different wall configurations were exposed to the entrance of a radiation furnace. Each configuration consisted of a gypsum board with PCM (exposed layer), a standard gypsum board, an EPS layer, a thermal insulation render layer with EPS (unexposed layer) and an insulation layer between the gypsum board with the PCM and the standard gypsum board. The insulation layers that were used for each configuration were: cavity (no insulation), mineral wool, EPS (conventional insulation materials) and VIP (super insulation material). Additional DSC measurements for all the examined materials were performed at inert (N2) and oxidized (air) atmosphere in order to support the analysis and findings of the furnace experiments. The DSC results revealed the well-known fire retardant behaviour of the GBs due to the dehydration process. The organic compounds incorporated in the PCM-GB (paraffin wax/PMMA), in the MW (resin binder), in the EPS (polymer), in the VIP (cellulose fiber) and in the TIR were volatilized (at inert atmosphere) and oxidized (at oxidized atmosphere) at temperatures higher than 300 °C. Regarding the “small-scale” furnace experiments, the results revealed, apart from the well-known fire retardant behaviour of the GBs, significant observations and conclusions for the fire behaviour and interaction of the utilized materials. The configuration with the EPS insulation layer performed almost the same as the configuration with the cavity (no insulation material). The latter was owed to the fact that the EPS melted at temperatures above 200 °C. The mineral wool insulation layer delayed the heat penetration through the assembly more than the cavity or the EPS until 500 °C where it started to melt. The utilization of a VIP insulation layer enhanced the fire resistance of the examined configuration when compared to the conventional insulation materials. It increased the time-to-failure of the configuration by an average of 68% when compared to the cavity (no insulation) and EPS case and 42% when compared to the mineral wool case. Overall, the work demonstrated that VIPs offer not only excellent thermal but also fire performance. The examined multilayer drywall assemblies can form the core part of lightweight dry wall systems, with very good fire performance.
Acknowledgments The authors acknowledge the financial support of the European Commission in the frame of the FP7 NMP project ‘ELISSA’, Grant agreement no: 609086.
References [1] IEA, Energy efficiency requirements in building codes, energy efficiency policies for new buildings, OECD/IEA, France, 2008. [2] R. Mateus, S. Neiva, L. Bragança, P. Mendonça, M. Macieira, Sustainability assessment of an innovative lightweight building technology for partition walls – comparison with conventional technologies, Build. Environ. 67 (2013) 147–159.
16
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
[3] M. Karimpour, M. Belusko, K. Xing, F. Bruno, Minimising the life cycle energy of buildings: review and analysis, Build. Environ. 73 (2014) 106–114. [4] M. Ahmad, A. Bontemps, H. Sallée, D. Quenard, Thermal testing and numerical simulation of a prototype cell using light wallboards coupling vacuum isolation panels and phase change material, Energy Build. 38 (2006) 673–681. [5] K. Peippo, P. Kauranen, P.D. Lund, A multicomponent PCM wall optimized for passive solar heating, Energy Build. 17 (4) (1991) 259–270. [6] H. Ge, V.R. McClung, S. Zhang, Impact of balcony thermal bridges on the overall thermal performance of multi-unit residential buildings: a case study, Energy Build. 60 (2013) 163–173. [7] N. Zhu, P. Hu, L. Xu, Z. Jiang, F. Lei, Recent research and application of ground source heat pump integrated with thermal energy storage systems: a review, Appl. Therm. Eng. 71 (2014) 142–151. [8] D. Patteeuw, K. Bruninx, A. Arteconi, E. Delarue, W. D’haeseleer, L. Helsen, Integrated modeling of active demand response with electric heating systems coupled to thermal energy storage systems, Appl. Energy 151 (2015) 306–319. [9] L. Navarro, A. de Garcia, Colclough, M. Browne, S.J. McCormack, P. Griffiths, L. F. Cabeza, Thermal energy storage in building integrated thermal systems: a review. Part 1. active storage systems, Renew. Energy 88 (2016) 526–547. [10] L. Navarro, A. de Garcia, Colclough, M. Browne, S.J. McCormack, P. Griffiths, L. F. Cabeza, Thermal energy storage in building integrated thermal systems: a review. Part 2. Integration as passive system, Renew. Energy 85 (2016) 1334–1356. [11] M. Alizadeh, S.M. Sadrameli, Development of free cooling based ventilation technology for buildings: thermal energy storage (TES) unit, performance enhacement techniques and design considerations – a review, Renew. Sustain. Energy Rev. 58 (2016) 619–645. [12] G.L. Kyriakopoulos, G. Arabatzis, Electrical energy storage systems in electricity generation: energy policies, innovative technologies, and regulatory regimes, Renew. Sustain. Energy Rev. 56 (2016) 1044–1067. [13] M. Hunger, A.G. Entrop, I. Mandilaras, H.J.H. Brouwers, M. Founti, The behavior of self-compacting concrete containing micro-encapsulated phase change materials, Cement Concr. Compos. 31 (10) (2009) 731–743. [14] I. Mandilaras, M. Stamatiadou, D. Katsourinis, G. Zannis, M. Founti, Experimental thermal characterization of a Mediterranean residential building with PCM gypsum board walls, Build. Environ. 61 (2013) 93–103. [15] M.D. Romero-Sánchez, C. Guillem-López, A.M. López-Buendía, M. Stamatiadou, I. Mandilaras, D. Katsourinis, M. Founti, Treatment of natural stones with phase change materials: experiments and computational approaches, Appl. Therm. Eng. 48 (2012) 136–143. [16] I.D. Mandilaras, D.A. Kontogeorgos, M.A. Founti, A hybrid methodology for the determination of the effective heat capacity of PCM enhanced building components, Renew. Energy 76 (2015) 790–804. [17] H. Kaasinen, The absorption of phase change substances into commonly used building materials, Sol. Energy Mater. Sol. Cells 27 (2) (1992) 173–179. [18] A. Pasupathy, R. Velraj, Effect of double layer phase change material in building roof for year round thermal management, Energy Build. 40 (3) (2008) 193–203. [19] C. Voelker, O. Kornadt, M. Ostry, Temperature reduction due to the application of phase change materials, Energy Build. 40 (5) (2008) 937–944. [20] D.A. Kontogeorgos, I.D. Mandilaras, M.A. Founti, Fire behaviour of regular and latent heat storage gypsum boards, Fire Mater. 39 (5) (2015) 507–517. [21] IEA/ECBCS Annex 39, Vacuum Insulation in the Building Sector, Systems and Applications, 2005.
[22] S.E. Kalnaes, B.P. Jelle, Review: vacuum insulation panel products: a state-ofthe-art review and future research pathways, Appl. Energy 116 (2014) 355–375. [23] B.P. Jelle, Traditional, state-of-the-art and future thermal building insulation materials and solutions – properties, requirements and possibilities, Energy Build. 43 (10) (2011) 2549–2563. [24] S.A. Al-Sanea, M.F. Zedan, Effect of thermal bridges on transmission loads and thermal resistance of building walls under dynamic conditions, Appl. Energy 98 (2012) 584–593. [25] I.D. Mandilaras, D.A. Kontogeorgos, I.A. Atsonios, C. Stark, S. Vidi, D. Marity, D. Seiler, D. Herfurth, M.A. Founti, Thermal performance of a steel skeleton/ drywall building insulated with Vacuum Insulation Panels, in: Proceedings of 12th International Vacuum Insulation Symposium, Nanjing, China, 19–21 September, 2015. [26] Dimos A. Kontogeorgos, Ioannis A. Atsonios, Ioannis D. Mandilaras, Maria A. Founti, Numerical investigation of the effect of vacuum insulation panels on the thermal bridges of a lightweight drywall envelope, in: Proceedings of VII International Congress on Architectural Envelopes, San Sebastian-Donostia, Spain, 27–29 May, 2015. [27] H. Cheng, G.V. Hadjisophocleous, Dynamic modeling of fire spread in building, Fire Saf. J. 46 (4) (2011) 211–224. [28] J. Xin, C. Huang, Fire risk analysis of residential buildings based on scenario clusters and its application in fire risk management, Fire Saf. J. 62(A) (2013) 72–78. [29] A. Ciudad, A.M. Lacasta, L. Haurie, J. Formosa, J.M. Chimenos, Improvement of passive fire protection in a gypsum panel by adding inorganic fillers: experiment and theory, Appl. Therm. Eng. 31 (17–18) (2011) 3971–3978. [30] K.G. Wakili, E. Hugi, Four types of gypsum plaster boards and their thermophysical properties under fire condition, J. Fire Sci. 27 (1) (2009) 27–43. [31] D. Kontogeorgos, M. Founti, Numerical investigation of simultaneous heat and mass transfer mechanisms occurring in a gypsum board exposed to fire conditions, Appl. Therm. Eng. 30 (2010) 1461–1469. [32] D. Kontogeorgos, I. Mandilaras, M. Founti, Scrutinizing gypsum board thermal performance at dehydration temperatures, J. Fire Sci. 29 (2) (2011) 111–130. [33] ASTME E119-00a, Standard Test Methods for Fire Tests of Building Construction and Materials, 2000. [34] D.A. Kontogeorgos, M.A. Founti, Gypsum board reaction kinetics at elevated temperatures, Thermochim. Acta 529 (2012) 6–13. [35] D.A. Kontogeorgos, M.A. Founti, Gypsum board dehydration kinetics at autogenous water vapour partial pressure, Thermochim. Acta 545 (2012) 141–147. [36] R. Jansson, Measurement of thermal properties at elevated temperatures – Brandforsk project 328-031, SP Swedish National Testing and Research Institute, SP Report 2004:46, Borås, Sweden. [37] S. Mehta, S. Biederman, S. Shivkumar, Thermal degradation of foamed polystyrene, J. Mater. Sci. 30 (1995) 2944–2949. [38] W. Sha, E.A. O’ Neill, Z. Guo, Differential scanning calorimetry study of ordinary Portland cement, Cement Concr. Res. (1999) 1487–1489. [39] M.I.G. Miranda, C.I.D. Bica, S.M.B. Nachtigall, N. Rehman, S.M.L. Rosa, Kinetical thermal degradation study of maize straw and soybean hull celluloses by simultaneous DSC–TGA and MDSC techniques, Thermochim. Acta 565 (2013) 65–71. [40] EN 1995-1-2, Eurocode 5, Design of Timber Structures – Part 1-2: General – Structural Fire Design, 2004.
Contents lists available at ScienceDirect
Fire Safety Journal journal homepage: www.elsevier.com/locate/firesaf
Experimental investigation of the fire resistance of multi-layer drywall systems incorporating Vacuum Insulation Panels and Phase Change Materials Dimos A. Kontogeorgos n, Georgios K. Semitelos, Ioannis D. Mandilaras, Maria A. Founti National Technical University of Athens, School of Mechanical Engineering, Laboratory of Heterogeneous Mixtures & Combustion Systems, Heroon Polytechniou 9, Zografou Campus, Athens 15780, Greece
art ic l e i nf o
a b s t r a c t
Article history: Received 10 September 2015 Received in revised form 22 January 2016 Accepted 27 January 2016
This paper studies the fire resistance of innovative high thermally insulated multilayer drywall assemblies incorporating conventional insulation materials, Phase Change Materials (PCMs) and Vacuum Insulation Panels (VIPs). An experimental study was developed and implemented into two directions. In the first direction, four different multilayer drywall configurations were subjected to fire temperatures up to 900 °C from one side, while the other side was at ambient conditions. Each configuration consisted of a gypsum board with PCMs (PCM-GB), a standard gypsum board (S-GB), an Expanded Polystyrene (EPS) layer, a thermal insulation render containing EPS (TIR) and an insulation layer located between the PCM-GB and the S-GB. A different insulation layer was used for each configuration: cavity (no insulation), EPS, mineral wool (MW) (both conventional insulation materials) and Vacuum Insulation Panels (VIP) (super insulation material). In the second direction, Differential Scanning Calorimetry (DSC) measurements, at inert (nitrogen) and oxidized (air) environments, were performed for all the utilized materials. DSC results indicated that at temperatures up to 200 °C, the gypsum boards (both PCM-GB and S-GB) act as fire retardants because of the dehydration process. The paraffin and PMMA components of the PCM started to evaporate and oxidize at temperatures higher than 200 °C and up to 500 °C. The resin binder of the mineral wool started to volatilize and oxidize at 265 °C, while at 500 °C the mineral wool started to melt. The volatilization of the EPS started at 275 °C, while the full volatilization and oxidation took place at the temperature range between 420 °C and 550 °C. The chemically bound water of the TIR dehydrated at the temperature range between 50 °C and 200 °C, while the EPS contained in the TIR behaved similar as the EPS sample. Finally, the cellulose fibers contained in the VIP volatilized and oxidized at the temperature range between 320 °C and 480 °C. Furnace results confirmed the fire resistance behavior of the gypsum boards indicated by the dehydration “plateau”. The wall assembly with the EPS layer found to behave similar to the assembly with the cavity due to the fact that the EPS melted at temperatures near 200 °C. The wall assembly with the mineral wool delayed the temperature rise until 500 °C where it started to melt. The VIP layer found to significantly delay the penetration of the heat through the drywall configuration when compared to the other configurations. According to the failure criteria regarding excessive temperature rise on the ambient facing side of the wall, the VIP layer was found to increase the time-to-failure by approximately 68%, with respect to the assembly with the cavity. On the other hand, the respective time increase for the conventional insulation materials was 2% and 19% for the EPS and the MW, respectively. & 2016 Elsevier Ltd. All rights reserved.
Keywords: Fire resistance Fire behavior VIP PCM EPS Mineral wool Super insulation materials DSC
1. Introduction In the last two decades, the need for sustainable buildings that minimize the consumption of raw materials and energy [1] has n Correspondence to: Laboratory of Heterogeneous Mixtures and Combustion Systems, Thermal Engineering Section, School of Mechanical Engineering, National Technical University of Athens, Heroon Polytechniou 9, Polytechnioupoli-Zografou, Athens 15780, Greece. E-mail address: [email protected] (D.A. Kontogeorgos).
http://dx.doi.org/10.1016/j.firesaf.2016.01.012 0379-7112/& 2016 Elsevier Ltd. All rights reserved.
gradually shifted the focus of the building sector towards lightweight constructions and nearly Zero Energy Buildings (nZEB). Lightweight multi-level buildings based on Dry Wall Systems (DWS) provide a safe, fast and long lasting solution to housing, particularly in high risk areas, such as highly seismic areas [2,3]. On the other hand, the presence of low density materials and thin envelopes in the lightweight buildings reduces the thermal efficiency (low thermal mass and low insulation) of the buildings. The latter, creates a number of new challenges for the design of
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
Nomenclature Abbreviations DSC DWS EPS MW nZEB
Differential Scanning Calorimetry Dry Wall System Expanded Polystyrene Mineral Wool Nearly Zero Energy Building
lightweight buildings. An attractive solution to this challenge appears to be the efficient integration of Phase Change Materials (PCMs) and Super Insulation Materials (SIM) in the lightweight building envelope [4]. The PCMs, due to their high amount of latent heat, utilized for thermal energy storage in building elements, are able to cope with thermal and electricity peak loads [5]. The SIMs, due to their very low thermal conductivity, provide efficient wall insulation solutions, as well as efficient solutions to thermal bridging effects [6]. Hence, the incorporation of PCMs and SIMs in DWS can lead to sustainable and thermally efficient lightweight constructions. Thermal Energy Storage (TES) in general and PCMs in particular, have been a research topic for the last 20 years. TES is an area of international interest dealing with energy saving, efficient and rational use of available resources [7–12]. It provides solutions in very specific areas: time delay and available power between production or availability of energy and its consumption in receiving systems (solar energy, cogeneration, etc.); security of energy supply (hospitals, computer centers, etc.); thermal inertia and thermal protection of the buildings. The thermal inertia and thermal protection are the areas where the PCMs have achieved a relatively high market penetration. Applications include, among others, their incorporation in the core of building materials. Particularly in the building sector, PCMs have been successfully integrated in the building fabric either by directly incorporating PCM microcapsules into commonly used building materials, such as in concrete, gypsum boards and natural stones [13–15] or by using separate layers of shape stabilized PCMs in wall assemblies [16]. The utilization of PCMs could lead to reduction of power requirements of the heating and cooling equipment commonly used to secure continuous energy supply. In the building context, the incorporation of PCMs into the core of building materials is intended to increase the thermal storage capacity of the building element [17]. The technology takes advantage of the latent heat of the PCM during the solid–liquid change of state to stabilize the temperature of the material and reduce the heat losses/gains from the building to the environment [18,19]. However, PCMs are usually mixtures of hydrocarbon molecules, which at temperatures higher than 200 °C evaporate. The gas mixture produced is flammable and may affect the fire endurance of the building element [20]. SIMs and particularly the Vacuum Insulation Panels (VIP) have been developed in the last decade, mainly for use in appliances, such as refrigerators and deep-freezers [21,22]. The reason hereto is the significantly low thermal conductivity (5–7 mW/(m K)), a factor of five to eight times better than the conventional insulation materials (Z 30 mW/(m K)) [23]. In the frame of building constructions, VIPs enable thin and highly insulation constructions to be realized for walls, floors and roofs. The motivation for examining the SIMs in buildings comes mainly from the difficulties involved in renovation, aesthetic considerations, as well as from the necessity to reduce the overall energy consumption of the buildings. The latter is accomplished by increasing the thermal resistance of the envelope's walls, as well as by reducing the effect
PCM PCM-GB PFP S-GB SIM TES TIR VIP
9
Phase Change Materials Gypsum board with Phase Change Materials Passive Fire Protection Standard gypsum board Super Insulation Material Thermal Energy Storage Thermal Insulation Render Vacuum Insulation Panel
of thermal bridges introduced when materials with higher thermal conductivity (e.g. steel studs) break through or partially break through a layer with higher thermal resistance (i.e. lower thermal conductivity) creating pathways of lower thermal resistance causing increased heat losses [24–26]. VIPs consist of inorganic compounds and thus, it is expected that they would enhance the fire resistance of a wall configuration. To the authors' knowledge, there is not any research found in the literature on the fire behavior of VIPs. Fire protection in the building technology is still extremely topical, especially due to the increase of the legislation and fire standard restrictions [27,28]. Dry wall systems (e.g. gypsum boards, cement boards etc.) are the most common systems used in lightweight steel skeleton buildings as Passive Fire Protection (PFP) systems for the steel structure (e.g. frames, studs etc.) [29]. This is due to the fact that they have very good fire behavior, associated with the water contained in their crystal structure, which is evaporated under fire conditions absorbing significant heat quantities from the fire and thus, delaying the heat penetration through the assembly [30–32]. The objective of the current work is to study for the first time the fire resistance of innovative and high thermally insulated wall assemblies in terms of energy storage and super insulation capabilities. The examined combinations of multilayer assemblies can find direct application in drywall construction due to their exceptionally good thermal insulation properties and relatively low thickness. An experimental study was developed and implemented, where “small-scale” furnace tests, as well as DSC measurements were performed in order to deeply analyze the fire resistance of the examined configurations. Within the furnace tests, four different multi-layer drywall configurations incorporating PCMs, VIPs and conventional insulation materials (mineral wool and expanded polystyrene) were subjected to fire temperatures up to 900 °C from one side, while the other side was at ambient conditions. Additional DSC measurements for all the components of the examined assemblies were performed in order to assess their thermal degradation at temperatures up to 600 °C. The main innovation of this work is owed to the study of the fire behavior of VIPs and PCMs, the interaction between different building materials that form a wall assembly under fire conditions and to the provision of experimental data in order to calibrate and validate dedicated fire numerical models.
2. Experimental study The overall experimental strategy followed two directions. Firstly, the chemical reactivity of the examined materials under inert and oxidized conditions was examined by performing DSC measurements. Secondly, “small-scale” furnace experiments were performed, in which multi-layer configurations, composed of different building materials, were positioned in front of a high temperature radiation furnace. The temperature evolution at different positions through the configurations was recorded, and together
10
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
Table 1 General information about the utilized materials. Material
k (W m
S-GB PCM-GB Mineral Wool EPS VIP
0.25 0.28 0.035 0.032 0.007
1
K
1
)
ρ (kg m
3
)
820 790 35 40 195
Cp (J kg
1
K
1
)
Composition – 19.1% encapsulated PCM (78% n-heptadecane with melting point near 22 °C, 22% PMMA) – – 98% fumed silica 2% cellulose fibers, organic binders and graphite based opacifiers
1000 1000 850 1300 800
subtracted from the DSC curve measured by having placed a sample mass in the same pan. In order to define the baseline at the temperature range where each reaction takes place, integral tangential baselines, provided from the apparatus software, were utilized. Finally, the average mass of the examined samples used in the DSC measurements was near 20 mg for the gypsum boards and the thermal insulation render and around 8 mg for the insulation materials (mineral wool, EPS and VIP). Each DSC measurement was repeated three times and the reproducibility error found to be within the instrument limits (i.e. 72%). 2.3. Furnace experiments
Fig. 1. Schematic diagram of the multi-layer wall configurations.
with the DSC results, their fire response and fire resistance was assessed. 2.1. Materials The materials that were used for all the experiments (DSC and “small-scale” furnace experiments) were commercially available. Table 1 summarizes the main general information of the utilized materials. 2.2. DSC experiments The chemical reactivity of the examined samples was defined by DSC measurements in a Stare SW 8.10, Mettler Toledo apparatus with an accuracy of 72%. The measurements were performed at the temperature range between 25 °C and 600 °C, using 10 °C/min and 20 °C/min heating rates, in inert (nitrogen) and oxidized (air) atmospheres, with a gas flow of 100 ml/min, using 20 μl aluminum crucibles. The apparatus was firstly calibrated at 156.6 °C using indium and at 419.6 °C using zinc, while before each measurement a blank DSC run was performed with an empty pan using the same conditions. The resulting blank curve was
The second part of the experimental study comprised four different “small-scale” furnace experiments. The experiments were designed in order to study the behavior of different building materials used in DWS and their interaction when exposed to fire conditions, as well as to serve as benchmark regarding the verification of detailed numerical fire models. More specifically, four different multilayer wall configurations, each consisting of five layers, were subjected to fire conditions from one side (i.e. furnace entrance), while the other side was at ambient conditions. According to standard ASTM E 119-00a [33] there is a degree of repeatability and reproducibility of fire testing. However, results depend on several factors, such as the type of assembly and material being tested, the nature of the boundary conditions and the details of the workmanship during the preparation of the specimens. For that reasons, it is impossible to quantify the anticipated error and thus, each of the “small-scale” furnace experiments was performed once. 2.3.1. Test specimens Fig. 1 illustrates a schematic diagram of the multi-layer wall configurations. As it is shown, beginning from the left side of the wall assembly, which is the layer that faces the furnace's entrance, the first layer is a gypsum board with PCMs (PCM-GB), the third layer is a standard gypsum board (S-GB), the fourth layer is an EPS and the fifth layer is a thermal insulation render layer containing EPS (TIR). The only difference between the experiments was that the second layer was different for each experiment; there was a cavity (no insulation) in experiment 1, an EPS in experiment 2, a Mineral Wool (MW) in experiment 3 and a VIP in experiment 4. The total thickness of all the configurations examined was 50 mm. Table 2 summarizes the list of the “small-scale” furnace experiments, as well as the materials used in each configuration and their thicknesses.
Table 2 List of performed experiments and utilized materials. Exp. number
Layer 1
Experiment Experiment Experiment Experiment
PCM-GB PCM-GB PCM-GB PCM-GB
1 2 3 4
(12.5 mm) (12.5 mm) (12.5 mm) (12.5 mm)
Layer 2
Layer 3
Layer 4
Layer 5
Cavity (12.5 mm) EPS (12.0 mm) MW (12.5 mm) VIP (10.0 mm)
S-GB S-GB S-GB S-GB
EPS EPS EPS EPS
TIR TIR TIR TIR
(12.5 mm) (12.5 mm) (12.5 mm) (12.5 mm)
(6.0 mm) (6.0 mm) (6.0 mm) (6.0 mm)
(6.5 mm) (7.0 mm) (6.5 mm) (6.5 mm)
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
Fig. 2. Specimen holder for the “small-scale” furnace experiments including the special ceramic insulation and the metal frame.
2.3.2. Specimen preparation As mentioned above, the only difference between each “smallscale” experiment was that the second layer, beginning from the layer exposed to the furnace entrance, was different for each experiment, while all the other layers were the same. The main idea of the “small-scale” experiments was the placement of each specimen in front of the entrance of a furnace. In order to fix each specimen in front of the furnace, it was positioned within a special specimen holder, which was consisted of a special ceramic insulation for high temperatures held by a metal frame, as shown in Fig. 2. The ceramic insulation was used in order to ensure the one
11
dimensionality of the heat transfer and minimize the lateral thermal losses. Fig. 3 shows how the materials, equipped with thermocouple sensors, were positioned inside the ceramic insulation The small gaps that occurred between the ceramic insulation and each material were filled in by means of a special ceramic paste that withstands temperatures up to 1300 °C. Each specimen was equipped with twenty K-type thermocouples with an accuracy of 70.4%, allowing the measurement of local temperatures at different locations within the specimen and on its surfaces, as well as of the surrounding air inside and outside the furnace. The thermocouples with the minimum possible diameter that withstand the temperature field at each position were chosen. The thermocouple sensors used for measuring the temperature evolution inside the furnace, on the surfaces of the PCM-GB and on the hot surface of the S-GB were thick (0.24 mm diameter) in order to withstand the high temperatures. On the other hand, the thermocouple sensors used for measuring the temperature evolution at the ambient and on the remaining surfaces were fine (0.12 mm diameter). The measuring tips of the thermocouple sensors within the wall assembly were placed in grooves less than 0.5 mm deep at the surfaces of the gypsum boards (both PCM-GB and S-GB), the EPS and the TIR. The good thermal contact of the tips with the materials during the measuring period was assured with the use of a special fire resistant adhesive (see Fig. 3). The thin adhesive layer, which covered the tips of the thermocouple sensors, was prevented the exposure of the tips, located on the exposed surface, from the furnace's radiation. Thus, no temperature correction for radiation losses was needed. Moreover, due to the fact that the radiation properties of the adhesive were similar to the respective radiation properties of the materials, the thermal radiation load on the exposed and unexposed surfaces of the wall configuration was
Fig. 3. Materials equipped with thermocouples positioned inside the ceramic insulation.
12
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
Fig. 4. Schematic diagram for the location of the thermocouples for each interface of the multi-layer wall configuration.
not affected by the thin adhesive layer. Fig. 4 visualizes the location of the thermocouples for each interface between the materials, which composed the specimen. For the temperature measurements, an Agilent data acquisition unit with an accuracy of 72% was utilized, while the temperatures were recorded and stored in intervals of 2s using the ‘LabVIEW’ software. 2.3.3. Experimental methodology Each specimen (i.e. multilayer wall configuration and specimen holder) was positioned in front of a Nabertherms B180 radiation furnace (250 mm height, 320 mm width and 550 mm depth), in such a way so that one surface of the specimen (PCM-GB) to be exposed to the furnace's thermal load and the other to be at ambient conditions. After the placement of the specimen in front of the furnace, the furnace was turned on, and the temperature value of 900 °C was set as the target temperature. The experiment was terminated when the temperature values recorded by the thermocouples were constant, reaching steady state conditions. Fig. 5 illustrates the overall experimental set-up and the specimen positioned in front of the furnace.
3. Observations, results and discussion 3.1. DSC experiments Fig. 6 illustrates the DSC results for all the examined materials at inert (N2) and oxidized (Air) environments. The respective
Fig. 5. Experimental set-up and specimen (including the specimen holder) positioned in front of the furnace.
results for the PCM-GB samples are shown in Fig. 6a. The typical GB dehydration characteristic peaks (1A and 1B) are observed at the temperature range between 100 °C and 200 °C, regardless the utilized atmosphere [34,35]. This process is endothermic, absorbs significant amounts of energy and it is thus, capable of slowing down the fire spread through the examined materials. Phenomena associated with the PCM compounds are also observed. The DSC
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
13
Fig. 6. DSC measurements of the core material at inert and oxidizes environments: (a) gypsum board with PCMs (PCM-GB), (b) standard gypsum board (S-GB), (c) mineral wool (MW), (d) expanded polystyrene (EPS), (e) thermal insulation render (TIR) and (f) Vacuum Insulation Panel (VIP).
analysis at inert and oxidized environment revealed the effect of PCM incorporation into the samples. At inert environment, the paraffin and PMMA, which compose the PCM, evaporate in the temperature range between 200 °C and 500 °C, as indicated by peaks 4A and 4B, while under oxidized environment the paraffin and PMMA oxidize in the same temperature range, as indicated by peaks 3A and 3B. Fig. 6b illustrates the DSC results of the S-GB samples. Once again, the typical dehydration process (peaks 1A and 1B) is observed regardless the utilized atmosphere. At temperatures near 380 °C, a small exothermic peak is observed, which corresponds to the reorganization of the GB's crystal mesh (peak 2) [34,35]. The latter exists but not observed in Fig. 6a at the oxidized atmosphere due to the presence of the exothermic behaviour of the PCM's paraffin and the PMMA at this temperature range. Fig. 6c illustrates the DSC results for the mineral wool samples. At inert conditions, a slightly endothermic reaction is observed, as indicated by the peak shown at 265 °C. This reaction is associated to the volatilization of the resin binder inside the mineral wool. On the other hand, at oxidized conditions the volatiles of this resin binder are oxidized, as indicated by peak 5 in Fig. 6c. Finally, temperature 500 °C is the onset temperature where the material begins to melt [36]. Fig. 6d illustrates the DSC results for the EPS. The endothermic reaction observed at inert environment (peak 6A) is owed to the break of the chemical bonds inside the EPS [37]. When the samples are heated at temperatures around 160 °C, the sample begins to melt. Under inert conditions the evaporation of the polymer starts at around 275 °C, while at temperatures between 420 °C and 450 °C the polymer has almost completely evaporated (peak 6A). The released volatiles oxidize in the presence of air (peak 7A). It should be noted that the released energy due to the oxidation of
the EPS at the temperature range between 275 °C and 450 °C is less than the absorbed energy due to the evaporation of the EPS and thus, the oxidation process is still an “endothermic” process at this temperature range (peak 7A). At the temperature range between 450 °C and 550 °C the oxidation of the EPS is clearly an exothermic process, as shown from peak 7B. Fig. 6e illustrates the DSC results for the thermal insulation render. The examined render is a mixture of cement and EPS. Thus, at the temperature range between 50 °C and 200 °C the endothermic dehydration process of the chemically bound water of the cement is observed (peak 9) [38]. The peaks shown near 300 °C and 450 °C are associated to the evaporation of the EPS (peaks 6A and 6B) and to the oxidation of the EPS’s volatiles (peaks 8A and 7B). Finally, the DSC results for the VIP samples (i.e. core material) are shown in Fig. 6f. As it is displayed, an endothermic process is observed at the temperature range between 20 °C and 100 °C (peak 10) regardless the utilized atmosphere, which is related to the loss of water, which is contained into the cellulose fibers [39]. Moreover, at temperatures near 320 °C and 480 °C, respectively, two peaks are observed at both inert and oxidized atmospheres. Peaks 11A and 11B are associated to the volatilization of the cellulose fibers, which is an endothermic process, while peaks 12A and 12B are associated to the oxidation of the cellulose fibers volatiles [39]. 3.2. Furnace experiments Fig. 7 shows the temperature evolution on the exposed surface, at the interfaces and on the unexposed surface of the furnace experiments, as well as the temperature evolution inside the furnace and at the ambient, as measured by the installed
14
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
Fig. 7. Temperature evolution: (a) inside the furnace, (b) on the exposed side, (c) at the interface between the PCM-GB and the insulation layer, (d) at the interface between the insulation layer and the S-GB, (e) at the interface between the S-GB and the EPS, (f) at the interface between the EPS and the TIR, (g) on the unexposed side and (h) at the ambient.
thermocouples. As mentioned above, the measuring tip of the thermocouple used for measuring the temperature evolution inside the furnace was thick in order to withstand the high temperatures. Therefore, the response time was high and attenuated all possible fluctuation (if any) of the temperature. On the other hand, the thermocouple used for measuring the temperature evolution of the ambient air was thin with very fast response. Therefore, what it appears as noise in Fig. 7h is actually the fluctuation of temperature due to the movement of the air near the surface of the wall assembly caused by the buoyant forces. It is clearly shown that the dehydration process of both PCMGB and S-GB significantly delays the temperature increase through the assembly's thickness. The latter is indicated by the observed temperature “plateau” shown at the interfaces between the different layers (Fig. 7c–g), and is more evident in the unexposed side of the assembly (Fig. 7g), where the temperature rise reduces considerably for a major period of time (∼30 min for the cavity case up to ∼70 min for the VIP case). When the dehydration process ends, there is a temperature increase at the positions behind the S-GB and up to the unexposed side (Fig. 7e–g). This can be explained by the fact that during the dehydration process, the dehydrated layers, behind the dehydration front, have significantly higher temperature than the non-dehydrated layers. Thus, when the dehydration effect ends, the low temperature layers, protected by the dehydration front, suddenly face significantly increased temperatures. The observed peaks shown in Fig. 7e–g are associated to the oxidation of the EPS volatiles, which takes place in the region near the EPS and TIR layers in the temperature range between 300 °C and 550 °C, in line with the DSC results (Fig. 6d and e). Thus, during the specific experiments, performed under ambient
oxidized conditions, the EPS's volatiles oxidized increasing the temperature recorded in the thermocouples. This temperature increase affects the temperature evolution at the interfaces in front of the EPS layer, as clearly shown by the temperature peaks in Fig. 7b–d. Finally, there was no indication of the PCM oxidation, as showed in the DSC results under oxidized conditions, due to the under-ventilated conditions inside the radiation furnace. The specimen with the VIP in the second layer (experiment 4) has a significantly better fire performance than the other assemblies with conventional insulation materials or with the cavity. The utilization of a VIP layer leads to a major delay of the heat penetration through the materials located between the VIP and up to the unexposed side of the assembly. This is clearly shown in Fig. 7e–g where the temperature “plateau” of experiment 4 is much more pronounced than the respective “plateau” observed in the other experiments. On the other hand, the conventional insulation materials utilized in the second layer (i.e. mineral wool and EPS), are slightly protecting the wall assembly. Regarding the EPS insulation case (experiment 2), due to the fact that it melts at around 200 °C, the specimen behaves almost the same as the specimen with the cavity (experiment 1), as shown in Fig. 7c–g. This means that the EPS does not offer any fire protection to the wall assembly. Regarding the Mineral Wool (experiment 3), due to the fact that its components melt near 500 °C, the wall assembly is protected until this temperature, as depicted from Fig. 7d–g, where the observed temperature “plateau” is maintained for a longer period of time than the respective “plateau” of experiments 1 (cavity) and 2 (EPS). The above observations are also shown in Fig. 8, which illustrates the infrared images of the unexposed side for all the “smallscale” experiments at different time intervals. As it is shown, there
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
15
reaches approximately 68% (with respect to the case with a cavity but no insulation material), while the respective increase percentages for the conventional insulation materials are near 2% for the EPS and 19% for the Mineral Wool.
4. Conclusions
Fig. 8. Infrared images of the unexposed side of all the examined configurations at 30 min time intervals.
Table 3 Time-to-failure values for each experiment. Experiment Experiment Experiment Experiment Experiment
1 2 3 4
Time-to-failure (min)
Increase (%)
80.1 81.6 95.2 134.9
– 2 19 68
is a small error introduced (estimated ∼5%) in the central part of the assembly due to the lateral heat from the ceramic insulation of the special holder. The latter proves the one dimensionality of the performed experiments and certifies the applicability of the measurements for validating numerical fire models. Regarding the behavior of the materials, it can be seen that the specimen with the VIP (experiment 4) delays significantly the heat penetration through the assembly. The latter is more obvious from the infrared images and the temperature distribution on the unexposed surface at 90 min, 120 min, 150 min and 180 min. The specimen with the EPS (experiment 2) behaves almost the same as the specimen with the cavity (experiment 1), as shown by the temperature distribution on the unexposed surfaces. Finally, the specimen with the mineral wool (experiment 3) behaves better that the specimens with the cavity (experiment 1) and the EPS (experiment 2), but not better than the specimen with the VIP (experiment 4). The timeto-failure values were calculated for each experiment, according to the failure criteria of the excessive temperature rise on the ambient facing side of the wall [40], which is valid when the temperature rise on the ambient facing side is a maximum of 180 °C or an average rise of 140 °C. The calculated results are tabulated in Table 3. It is obvious that the utilization of a VIP instead of a conventional insulation material leads to a significant increase of the time-to-failure. In this case, the time-to-failure increase
In this paper, an experimental study was developed and implemented in order to investigate the fire resistance and behaviour of innovative, high thermally insulated, multilayer drywall assemblies incorporating conventional insulation materials, VIPs and PCMs when exposed to fire conditions. Four different wall configurations were exposed to the entrance of a radiation furnace. Each configuration consisted of a gypsum board with PCM (exposed layer), a standard gypsum board, an EPS layer, a thermal insulation render layer with EPS (unexposed layer) and an insulation layer between the gypsum board with the PCM and the standard gypsum board. The insulation layers that were used for each configuration were: cavity (no insulation), mineral wool, EPS (conventional insulation materials) and VIP (super insulation material). Additional DSC measurements for all the examined materials were performed at inert (N2) and oxidized (air) atmosphere in order to support the analysis and findings of the furnace experiments. The DSC results revealed the well-known fire retardant behaviour of the GBs due to the dehydration process. The organic compounds incorporated in the PCM-GB (paraffin wax/PMMA), in the MW (resin binder), in the EPS (polymer), in the VIP (cellulose fiber) and in the TIR were volatilized (at inert atmosphere) and oxidized (at oxidized atmosphere) at temperatures higher than 300 °C. Regarding the “small-scale” furnace experiments, the results revealed, apart from the well-known fire retardant behaviour of the GBs, significant observations and conclusions for the fire behaviour and interaction of the utilized materials. The configuration with the EPS insulation layer performed almost the same as the configuration with the cavity (no insulation material). The latter was owed to the fact that the EPS melted at temperatures above 200 °C. The mineral wool insulation layer delayed the heat penetration through the assembly more than the cavity or the EPS until 500 °C where it started to melt. The utilization of a VIP insulation layer enhanced the fire resistance of the examined configuration when compared to the conventional insulation materials. It increased the time-to-failure of the configuration by an average of 68% when compared to the cavity (no insulation) and EPS case and 42% when compared to the mineral wool case. Overall, the work demonstrated that VIPs offer not only excellent thermal but also fire performance. The examined multilayer drywall assemblies can form the core part of lightweight dry wall systems, with very good fire performance.
Acknowledgments The authors acknowledge the financial support of the European Commission in the frame of the FP7 NMP project ‘ELISSA’, Grant agreement no: 609086.
References [1] IEA, Energy efficiency requirements in building codes, energy efficiency policies for new buildings, OECD/IEA, France, 2008. [2] R. Mateus, S. Neiva, L. Bragança, P. Mendonça, M. Macieira, Sustainability assessment of an innovative lightweight building technology for partition walls – comparison with conventional technologies, Build. Environ. 67 (2013) 147–159.
16
D.A. Kontogeorgos et al. / Fire Safety Journal 81 (2016) 8–16
[3] M. Karimpour, M. Belusko, K. Xing, F. Bruno, Minimising the life cycle energy of buildings: review and analysis, Build. Environ. 73 (2014) 106–114. [4] M. Ahmad, A. Bontemps, H. Sallée, D. Quenard, Thermal testing and numerical simulation of a prototype cell using light wallboards coupling vacuum isolation panels and phase change material, Energy Build. 38 (2006) 673–681. [5] K. Peippo, P. Kauranen, P.D. Lund, A multicomponent PCM wall optimized for passive solar heating, Energy Build. 17 (4) (1991) 259–270. [6] H. Ge, V.R. McClung, S. Zhang, Impact of balcony thermal bridges on the overall thermal performance of multi-unit residential buildings: a case study, Energy Build. 60 (2013) 163–173. [7] N. Zhu, P. Hu, L. Xu, Z. Jiang, F. Lei, Recent research and application of ground source heat pump integrated with thermal energy storage systems: a review, Appl. Therm. Eng. 71 (2014) 142–151. [8] D. Patteeuw, K. Bruninx, A. Arteconi, E. Delarue, W. D’haeseleer, L. Helsen, Integrated modeling of active demand response with electric heating systems coupled to thermal energy storage systems, Appl. Energy 151 (2015) 306–319. [9] L. Navarro, A. de Garcia, Colclough, M. Browne, S.J. McCormack, P. Griffiths, L. F. Cabeza, Thermal energy storage in building integrated thermal systems: a review. Part 1. active storage systems, Renew. Energy 88 (2016) 526–547. [10] L. Navarro, A. de Garcia, Colclough, M. Browne, S.J. McCormack, P. Griffiths, L. F. Cabeza, Thermal energy storage in building integrated thermal systems: a review. Part 2. Integration as passive system, Renew. Energy 85 (2016) 1334–1356. [11] M. Alizadeh, S.M. Sadrameli, Development of free cooling based ventilation technology for buildings: thermal energy storage (TES) unit, performance enhacement techniques and design considerations – a review, Renew. Sustain. Energy Rev. 58 (2016) 619–645. [12] G.L. Kyriakopoulos, G. Arabatzis, Electrical energy storage systems in electricity generation: energy policies, innovative technologies, and regulatory regimes, Renew. Sustain. Energy Rev. 56 (2016) 1044–1067. [13] M. Hunger, A.G. Entrop, I. Mandilaras, H.J.H. Brouwers, M. Founti, The behavior of self-compacting concrete containing micro-encapsulated phase change materials, Cement Concr. Compos. 31 (10) (2009) 731–743. [14] I. Mandilaras, M. Stamatiadou, D. Katsourinis, G. Zannis, M. Founti, Experimental thermal characterization of a Mediterranean residential building with PCM gypsum board walls, Build. Environ. 61 (2013) 93–103. [15] M.D. Romero-Sánchez, C. Guillem-López, A.M. López-Buendía, M. Stamatiadou, I. Mandilaras, D. Katsourinis, M. Founti, Treatment of natural stones with phase change materials: experiments and computational approaches, Appl. Therm. Eng. 48 (2012) 136–143. [16] I.D. Mandilaras, D.A. Kontogeorgos, M.A. Founti, A hybrid methodology for the determination of the effective heat capacity of PCM enhanced building components, Renew. Energy 76 (2015) 790–804. [17] H. Kaasinen, The absorption of phase change substances into commonly used building materials, Sol. Energy Mater. Sol. Cells 27 (2) (1992) 173–179. [18] A. Pasupathy, R. Velraj, Effect of double layer phase change material in building roof for year round thermal management, Energy Build. 40 (3) (2008) 193–203. [19] C. Voelker, O. Kornadt, M. Ostry, Temperature reduction due to the application of phase change materials, Energy Build. 40 (5) (2008) 937–944. [20] D.A. Kontogeorgos, I.D. Mandilaras, M.A. Founti, Fire behaviour of regular and latent heat storage gypsum boards, Fire Mater. 39 (5) (2015) 507–517. [21] IEA/ECBCS Annex 39, Vacuum Insulation in the Building Sector, Systems and Applications, 2005.
[22] S.E. Kalnaes, B.P. Jelle, Review: vacuum insulation panel products: a state-ofthe-art review and future research pathways, Appl. Energy 116 (2014) 355–375. [23] B.P. Jelle, Traditional, state-of-the-art and future thermal building insulation materials and solutions – properties, requirements and possibilities, Energy Build. 43 (10) (2011) 2549–2563. [24] S.A. Al-Sanea, M.F. Zedan, Effect of thermal bridges on transmission loads and thermal resistance of building walls under dynamic conditions, Appl. Energy 98 (2012) 584–593. [25] I.D. Mandilaras, D.A. Kontogeorgos, I.A. Atsonios, C. Stark, S. Vidi, D. Marity, D. Seiler, D. Herfurth, M.A. Founti, Thermal performance of a steel skeleton/ drywall building insulated with Vacuum Insulation Panels, in: Proceedings of 12th International Vacuum Insulation Symposium, Nanjing, China, 19–21 September, 2015. [26] Dimos A. Kontogeorgos, Ioannis A. Atsonios, Ioannis D. Mandilaras, Maria A. Founti, Numerical investigation of the effect of vacuum insulation panels on the thermal bridges of a lightweight drywall envelope, in: Proceedings of VII International Congress on Architectural Envelopes, San Sebastian-Donostia, Spain, 27–29 May, 2015. [27] H. Cheng, G.V. Hadjisophocleous, Dynamic modeling of fire spread in building, Fire Saf. J. 46 (4) (2011) 211–224. [28] J. Xin, C. Huang, Fire risk analysis of residential buildings based on scenario clusters and its application in fire risk management, Fire Saf. J. 62(A) (2013) 72–78. [29] A. Ciudad, A.M. Lacasta, L. Haurie, J. Formosa, J.M. Chimenos, Improvement of passive fire protection in a gypsum panel by adding inorganic fillers: experiment and theory, Appl. Therm. Eng. 31 (17–18) (2011) 3971–3978. [30] K.G. Wakili, E. Hugi, Four types of gypsum plaster boards and their thermophysical properties under fire condition, J. Fire Sci. 27 (1) (2009) 27–43. [31] D. Kontogeorgos, M. Founti, Numerical investigation of simultaneous heat and mass transfer mechanisms occurring in a gypsum board exposed to fire conditions, Appl. Therm. Eng. 30 (2010) 1461–1469. [32] D. Kontogeorgos, I. Mandilaras, M. Founti, Scrutinizing gypsum board thermal performance at dehydration temperatures, J. Fire Sci. 29 (2) (2011) 111–130. [33] ASTME E119-00a, Standard Test Methods for Fire Tests of Building Construction and Materials, 2000. [34] D.A. Kontogeorgos, M.A. Founti, Gypsum board reaction kinetics at elevated temperatures, Thermochim. Acta 529 (2012) 6–13. [35] D.A. Kontogeorgos, M.A. Founti, Gypsum board dehydration kinetics at autogenous water vapour partial pressure, Thermochim. Acta 545 (2012) 141–147. [36] R. Jansson, Measurement of thermal properties at elevated temperatures – Brandforsk project 328-031, SP Swedish National Testing and Research Institute, SP Report 2004:46, Borås, Sweden. [37] S. Mehta, S. Biederman, S. Shivkumar, Thermal degradation of foamed polystyrene, J. Mater. Sci. 30 (1995) 2944–2949. [38] W. Sha, E.A. O’ Neill, Z. Guo, Differential scanning calorimetry study of ordinary Portland cement, Cement Concr. Res. (1999) 1487–1489. [39] M.I.G. Miranda, C.I.D. Bica, S.M.B. Nachtigall, N. Rehman, S.M.L. Rosa, Kinetical thermal degradation study of maize straw and soybean hull celluloses by simultaneous DSC–TGA and MDSC techniques, Thermochim. Acta 565 (2013) 65–71. [40] EN 1995-1-2, Eurocode 5, Design of Timber Structures – Part 1-2: General – Structural Fire Design, 2004.