*2.1. Materials*

In this experimental work, the three main components of materials used for the preparation of intumescent coatings were (1) the halogen-free flame retardant additives: Ammonium polyphosphate (particle size <15 μm), melamine (particle size <40 μm), pentaerythritol (particle size <40 μm), (2) the flame retardant fillers: Aluminum hydroxide with a specific surface area in a range between 0.5 to 50 m2/g, magnesium hydroxide (specific surface area <8 m2/g), titanium dioxide (particle size <40 μm, specific surface area = 150 m2/g) and eggshell bio-filler (mean particle size = 22.99 μm and specific surface area = 148.41 m2/g) [11] and (3) the polymer binder: Acrylic resin, which has slow-burning or even self-extinguishing behavior when exposed to fire. Moreover, it does not generate harmful smoke or gases. The eggshell powder preparation is shown in Figure 1.

**Figure 1.** Flow chart of chicken eggshell (ES) powder preparation.

Four kinds of thin-film intumescent coatings W, X, Y and Z were prepared and tested to evaluate the fire protection properties and combustion performances of the coatings. The sample details are tabulated in Table 1.


**Table 1.** Specifications of experimental materials.

\* Sample sized for cone calorimeter test = 10 cm (*l*) × 10 cm (*w*) × 1.5 mm (*t*)

### *2.2. Fire Protective Test*

The fire protective test allowed the observation of the development of the char layer and the evolution of temperature when exposed to fire to determine the performance of the intumescent fire protective coatings. The intumescent formulation was applied onto a steel plate after being grit-blasted (dimensions of steel plate: 100 <sup>×</sup> 100 <sup>×</sup> 2.6-mm3) by using a gun sprayer. This method was repeated 3–5 times until a 2.0 ± 0.2 mm coating thickness was attained. The Bunsen burner of the gas tank consumption was about 160 g/h, and the coated steel plate with the intumescent coating was mounted vertically and tested for 60 min of fire (about 1000 ◦C). In this fire test, 400 ◦C was chosen as a critical temperature for the coated steel under the small-scale fire test. The time-temperature curves of the coated steel plates were recorded and verified using a model of 307/308 hand-held mini thermometer. The temperature profile and thickness of the char layer on the backside of the steel plates were recorded and evaluated (see Figure 2).

**Figure 2.** Schematic of the experimental setup for fire protective test [31]. Figure 2 is adapted with permission from [31]. Copyright 2018 Elsevier.

#### *2.3. BS 476: Part 6 Fire Test*

The fire propagation test according to BS 476: Part 6 was conducted and evaluated in this experimental work [32]. The experimental method consisted of exposing the coated steel with an intumescent formulation to a standard small flame for 20 min, with 2 kW of an extra irradiance from the third to the final minute of the fire propagation test. The temperature of the grown ignition intumescent coatings was recorded. This was compared to the temperatures produced from the steel plate coated with the intumescent coating. The result stated was the index of fire propagation, which offers a relative measure of the involvement to the evolution of the fire made by the flat coating surface. The coated steel plate with the intumescent fire protective coating was exposed to the fire conditions. To be Class 0 certified the fire propagation index (*I*) must be below 12. In this fire test, a better fire protective material was determined by a lower numerical value of the index.

In addition, the heat rate and amount of heat grown by the coating sample were measured and considered under prearranged conditions. The steel plates with a standard dimension: <sup>225</sup> <sup>×</sup> <sup>225</sup> <sup>×</sup> 2.3-mm<sup>3</sup> were coated with a thickness of 1.5 <sup>±</sup> 0.1 mm intumescent coating. Moreover, the index of fire propagation performance was calculated using the below equations:

$$I\_1 = \sum\_{t=0.5}^{t=3} \frac{\theta\_m - \theta\_c}{10t} \tag{1}$$

$$I\_2 = \sum\_{t=4}^{t=10} \frac{\theta\_{\text{m}} - \theta\_{\text{c}}}{10t} \tag{2}$$

$$I\_3 = \sum\_{t=12}^{t=20} \frac{\theta\_m - \theta\_c}{10t} \tag{3}$$

$$I = I\_1 + I\_2 + I\_3 \tag{4}$$

Where:


#### *2.4. Sample Preparation for the Cone Calorimeter Test*

Before the cone calorimeter test, sample sizes (10 cm × 10 cm × 1.5 mm) were maintained at 50 ± 5% relative humidity (RH) and 23 ± 2 ◦C. The pretreated coating samples were enveloped with aluminum foil (thickness: 0.03–0.05 mm) with the shiny side of the aluminum foil facing the sample. The coating sample was wrapped without any treatment and the non-exposed surface was covered with foil, which typically forms to the cone calorimeter. One of the most acceptable and internationally recognized fire testing apparatuses is the cone calorimeter. This equipment test was conducted in accordance with the ISO 5660-1 standard. The cone calorimeter is used to examine the fire characteristics of the sample with different measurements simultaneously. The most important parameter to determine a fire's hazard level is to obtain the value of the heat release rate (HRR) of the sample. The specifications of the samples are shown in Table 1.

In this cone calorimeter test, the prepared coating sample and the holder were located on a mass measurement device. All the experiment works were evaluated by employing the coating samples in the same holder under a heater of the cone calorimeter. The fire situation comprised four stages: (1) Ignition, (2) growth, (3) fully developed, and (4) decay. The heat flux of 50 kW/m<sup>2</sup> was set and conducted by corresponding to the fully developed fire stage. The distance between the cone and the coating sample was 6 cm. The spark power and the igniter were removed when the ignition or

temporary flame occurred and the time was recorded. If the flame went out after removing the spark power, the igniter was re-inserted within 5 s and then the spark was maintained until test completion. The coating sample and sample holder were detached after collection of all data. Each pretreated coating samples were tested three times according to the standard. The experimental data of the coating sample were calculated based on the average of three tests [33].
