INVESTIGATION OF WEATHERING EFFECTS ON PERFORMANCE OF INTUMESCENT COATINGS USED IN FIRE PROTECTION OF WOOD STRUCTURES
Intumescent coating is one of many passive methods to protect wood structures in the Wildland-Urban Interface (WUI) from fire. \iffalse The mechanism of such coating can be explained in three stages. First, acid release reaction is triggered by the heat from a fire. Then, the carbonization process starts, and carbon as a protection layer forms. Finally, gas formation produces an expanded foam which isolates the substrate from the heat and oxygen.\fi The intumescent coating forms a thick layer of carbonaceous char that protects the substrates from a fire attack. However, when the intumescent coating is used on the exterior of a structure, weathering may affect the process of the intumescent mechanism and the effectiveness of the coating. Also, coating chemicals under weathering effects may change into more ignitable substances and cause fire damage to the protected wood structure. The purpose of this study was to investigate the weathering effect on the performance of the intumescent coating. There were two main concerns, the first concern was the possibility of decreasing ignition resistance of the coatings due to the weathering effect after a relatively short time. The second concern was the possibility of increasing the flammability of the coated specimens after weathering. To address these two concerns, the effects of weathering on the performance of fire retardant intumescent coating was studied. Calorimetry was used to collect the combustion data that formed 96 different classes of samples after various periods of natural weathering. The classes included samples that were exposed to the weathering effects for zero, three, six, and twelve months. Also, the samples were oriented northward and southward on the vertical weathering fence. Three different commercially available coating types were used to compare their performances to the weathering effects. Then, randomly selected specimens from each class were tested in the cone calorimeter at three different heat flux levels. The data consisted of the following measured combustion properties; Time to Ignition (TTI), Heat Release Rate (HRR), Mass Loss Rate (MLR), and Effective Heat of Combustion (EHC). To measure these combustion properties, the cone calorimeter was employed. TTI, HRR and MLR are more important to address the first concerns. They can play an important role in ignition phase and propagation phase of a fire. However, EHC is the primary parameter that can be used to address the second concern of this study. If the intumescent coating materials were changed to a more combustible fuel (with higher EHC) than uncoated wood by natural weathering, it would be more reasonable to avoid useing them as fire protection. In the next step, the ANOVA and MANOVA methods were employed as statistical tools to investigate the differences among the classes. Especially, the combustion properties were collected and compared for the 4 different weathering intervals to examine the effect of weathering on the two concerns. In addition, the machine learning method of Multi-class Decision Tree was employed to estimate the rank of each treatment effect on the response variables. The machine learning analysis showed that coating type is the most important factor in the ignition phase. TTI of samples was more affected by coating than the other involved parameters. However, in the propagation phase of fires, where HRR plays the major role, the weathering was the most effective factor. In the next step, the effects of weathering on all type of coated specimens were compared. In regards to the first concern: in almost all cases, the intumescent coatings showed a significant protection before weathering process. However, the fire protection decreased due to weathering, with an exception seen in the weathering effects on EHC. The weathering process decreased the EHC in almost all cases. The degradation of the coatings was different by type. In the next stage, the performance of coated specimens was compared to uncoated counterparts. This comparison showed the advantage of each type in different weathering time intervals. In some cases, there is no statistical evidence of different performance between coated specimens than uncoated specimens in 12 months of weathering. In addressing the second concern, the increasing combustibility was not seen in the weathering process of the intumescent coating except in one case of HRR after 12 months of weathering. The most important factor related to the second concern was that EHC did not increase by weathering. In addition, the results showed that the water-based intumescent coatings with thicker coated layers displayed overall better performance (type C and A). From this result it could demonstrate that the physical properties such as thickness of the coatings may affect the life time of the coatings.