Preparation and Performance Study of Flame-Retardant Composite Filling Materials for Tunnel Surrounding Rock

Abstract

Rigid polyurethane foam (RPUF) is a common filling material for tunnels surrounding rock in China. The Chinese national standard explicitly stipulates that RPUF, utilized as a tunnel filling material, must adhere to the following criteria: a thermal conductivity of ≤0.1 W/(m·K), a compressive strength of ≥150 kPa, a limiting oxygen index of ≥26%, and a flame-retardant grade of B2. However, the flame-retardant grade B2 is still possible to burn in the special environment of the tunnel. In view of the strict requirements of national standards for thermal conductivity, compressive strength, and flame-retardant performance of RPUF, this study focuses on optimizing the comprehensive performance of RPUF through scientific matching of flame retardants. The aim is to prepare RPUF that not only meets the national standard but also can reach B1 level. Three flame retardants, melamine polyphosphate (MPP), expandable graphite (EG), and low melting point glass powder (LGP), were selected for the ratio test. Measurement correlation coefficient. A comprehensive analysis of these test results was conducted. The results show that: When the overall proportion of EG-MPP-LGP is 41% and EG:MPP:LGP = 26:13:2. Its thermal conductivity is 0.0555 W/(m·K), compressive strength is 216.72 kPa, and the limiting oxygen index is 32.2%, which increases by 74% compared with pure RPUF. The flame-retardant grade achieved is B1, categorizing it as a flame-retardant material. Additionally, all other properties measured align with national standards. This innovative preparation method provides material support for tunnel safety engineering and has practical value.

1. Introduction

With the increase in railway construction, the number of tunnels also increases, and the demand for filling materials is increasing day by day as one of the essential materials in tunnel construction. However, the flame retardancy of RPUF is very poor, and the limiting oxygen index (LOI) is only 18.5%, which is a flammable material and easy to burn [1]. The tunnel environment poses unique challenges. In the event of the ignition of flammable materials, the fire spreads rapidly and intensifies swiftly, making it difficult to contain. The heat and smoke generated are not easily dispersed, leading to a rapid rise in temperature inside the tunnel. And it is hard to see what is going on outside the tunnel. Such an environment significantly hinders personnel evacuation and rescue efforts, ultimately resulting in severe consequences of accidents [2]. Therefore, in order to reduce the possibility of similar accidents and the severity of accidents, China’s “rigid polyurethane foam for building insulation” and “polyurethane rigid foam composite insulation board” and other relevant laws and regulations for the rigid polyurethane foam used in tunnels are clearly stipulated: The flame-retardant grade of RPUF for tunnel use should not be less than B2, and the LOI should be greater than 26%. Although the materials currently used in China can meet the limit oxygen index of more than 26%, the fire grade B2 is still flammable material. In the special environment of the tunnel, continuous high temperatures will also burn, making it easy to lead to the spread of fire. Therefore, it is necessary to improve the flame-retardant performance of RPUF.

In recent years, the research on improving the flame retardancy of RPUF has been widely concerned and reported in academia and industry. In some studies, the limiting oxygen index (LOI) of RPUF was successfully raised to more than 36%, and the flame retardancy was significantly enhanced. However, this achievement is not yet sufficient for direct application in the harsh environment of tunnels. The reason is that most of the existing studies focused on the single dimension optimization of flame-retardant performance. Therefore, a comprehensive evaluation of other key performance indicators, such as thermal insulation properties and mechanical properties, required for RPUF in tunnel applications is neglected. Such as thermal insulation properties and mechanical properties. In China, the “rigid polyurethane foam for building insulation” and “polyurethane rigid foam composite insulation board” and other relevant national standards have clear requirements for the thermal insulation and mechanical properties of tunnel RPUF: Thermal conductivity should be ≤0.1 W/(m·K), and compression strength should be ≥150 kPa. Therefore, the premise of the study is that the thermal insulation and mechanical properties of RPUF meet the national standards. The method of adding flame retardants in the process of preparing RPUF was selected to improve the flame retardancy of RPUF. The objective of this study is to prepare rigid polyurethane foam with thermal conductivity ≤ 0.1 W/(m·K), compressive strength ≥ 150 kPa, and LOI ≥ 32% by adding flame retardants.

In the preparation of RPUF, the choice of flame retardants is very important. The current mainstream flame-retardant types are roughly divided into four categories: halogen flame retardants, expansion flame retardants, phosphorus flame retardants, and ceramic flame retardants. Halogenated flame retardants produce gaseous hydrogen halide HX by thermal decomposition. At high temperatures, the hydrogen halide interacts with the active free radicals of the combustion reaction to terminate the chain reaction. At the same time, the acidity of gaseous hydrogen halide can promote the dehydration and carbonization of materials to form a fire-retardant carbon layer [3,4]. Therefore, adding halogenated flame retardants can make the flame retardancy of RPUF very good. However, halogen flame retardants will produce a lot of toxic gas when they burn. It has brought great harm to people’s physical safety and the environment. Therefore, its application under the condition of sustainable development strategy is limited [5]. In recent years, international environmental organizations have banned the use of halogenated flame retardants [6]. Therefore, the development and research of halogen-free flame-retardant rigid polyurethane foam materials is now a general trend [7]. Intumescent flame retardants, phosphorous flame retardants, and ceramic flame retardants are halogen-free flame retardants. In the selection of flame retardants, the test focuses on these three types of flame retardants.

Intumescent flame retardant is based on expandable graphite (EG). EG has been widely used in the preparation of RPUF in recent years, and its advantages and disadvantages are summarized below: Adding EG to RPUF can significantly improve the flame retardancy [8,9,10,11]. It has a maximum limiting oxygen index of 36%. EG is resource-rich, environmentally friendly, and matches the expansion temperature and RPUF decomposition temperature. The flame-retardant working range of EG is 200~280 °C [12,13]. The sulfuric acid in the interlayer lattice will undergo a REDOX reaction with the carbon atoms between the graphite layers to produce a large amount of gases such as SO₂, CO₂, and H₂O and heat [14]. The worm-like expanded carbon layer is formed on the surface of the material through its own physical expansion. However, because EG is an inorganic flame retardant, there will still be compatibility problems when it is applied to RPUF. Therefore, when EG is added to the polymer matrix, the poor compatibility between EG and the matrix leads to the deterioration of the mechanical properties of the composite. Wang Xinchao et al. [15] studied the effect of EG on the mechanical properties of RPUF. The results show that, when the amount of EG is 30%, the compressive strength decreases from 0.51 MPa to 0.29 MPa. Wang Shengcheng et al. [16,17] studied the effect of the combination of phosphate-based flame retardants and EG on the mechanical properties of rigid polyurethane. The results show that the prepared RPUF decreases linearly with the increase in the mass of expandable graphite. In summary, EG can significantly increase the limiting oxygen index of RPUF to more than 32%. This indicates the excellent flame-retardant properties of EG. However, with the increase in EG addition, the mechanical properties of RPUF may be negatively affected to some extent. In order to balance this contradiction, EG and phosphate-based flame retardants were mixed in the preparation of RPUF. It can not only maintain or even enhance the overall flame-retardant effect [18,19,20,21], but also effectively make up for the adverse effects of EG on mechanical properties when used alone and achieve complementary advantages in performance.

The main feature of phosphorous flame retardants is that they contain phosphate groups. On the one hand, the formation of a carbonization layer can prevent further pyrolysis of polymers. On the other hand, it can prevent the internal thermal decomposition products from entering the gas phase to participate in the combustion process [22]. The mainstream phosphorous flame retardants include melamine polyphosphate (MPP), red phosphorus, ammonium polyphosphate, dimethyl methylphosphonate, and diethyl aluminum hypophosphate. It can be seen from the literature that adding MPP to the preparation of RPUF can significantly increase its limiting oxygen index (LOI). It meets UL 94 V-0 standards. This indicates that the material exhibits excellent flame retardancy in combustion tests [23]. MPP is used in different fields and works well. In some fields, the limiting oxygen index of the material can reach more than 32% [24,25,26]. Additionally, the flame-retardant performance, the addition of MPP can also improve the mechanical properties of RPUF to a certain extent, including compressive strength and bending strength. This makes RPUF more stable and reliable in applications [27]. However, the addition of MPP will lead to the deterioration of insulation performance. Wang Zhengzhou et al. [28] found in the test that the thermal conductivity of flame-retardant RPUF supplemented with 15% MPP increased to 0.0259 W/(m·K). Although MPP can effectively improve the flame-retardant performance of RPUF, it may also bring about the side effect of increasing thermal conductivity. This makes the insulation performance of RPUF worse. In order to make up for this shortage, the introduction of ceramic flame-retardant materials has become a feasible solution. The material can optimize the insulation performance of RPUF while maintaining or enhancing flame-retardant properties.

Ceramic flame-retardant material is a new type of functional pollution-free flame-retardant material, also known as low melting point glass powder (LGP). LGP is often used as a binding material, sintered material, and high-temperature solvent to meet some requirements of processing production and scientific research [29,30]. According to published research data, when the addition of LGP is 7%, the thermal conductivity of polyurethane foam reaches a minimum of 0.0235 W/(m·K). This indicates that LGP can significantly reduce the thermal conductivity of rigid polyurethane foam with an appropriate addition amount [31]. By studying the mechanism of LGP, LGP can not only improve the flame-retardant performance of rigid polyurethane foam but also improve its thermal conductivity [32,33,34]. Therefore, LGP was selected to be added to the intumescent flame-retardant system. In order to achieve the double improvement of flame-retardant and heat preservation performance.

In summary, MPP, EG, and LGP are added in the process of preparing RPUF. By changing the ratio of three flame retardants, it is expected to prepare rigid polyurethane foam with thermal conductivity ≤ 0.1 W/(m·K), compressive strength ≥ 150 kPa, and LOI ≥ 32%.

2. Experiment

2.1. Experimental Material

In the experiment, flame retardants were added during the preparation of RPUF. The RPUF raw materials and flame retardants used are listed in

Table 1. Experimental materials.

RPUF Feedstock Reagent	Specification
Polyether polyol (PET-4110)	Industrial grade
4,4′-Diphenylmethane diisocyanate (MDI)	Industrial grade
n-Pentane	AR
Distilled water	AR
Flame Retardant	Specification
Melamine polyphosphate (MPP)	Industrial grade
Expandable graphite (EG)	Industrial grade
Low melting point glass powder (LGP)	Industrial grade

. Flame retardants are shown in Figure 1.

2.2. Experimental Ratio

The experiment is divided into five parts. Part A and Part B are the research on the performance of EG and MPP separately added to RPUF, and Part C–E is the matching test. The specific experimental arrangements are presented in

Table 2. Experimental Ratio.

Number	Amount of Flame Retardant Added			RPUF Raw Material Content (%) (100-Amount of Flame Retardant Added)
	EG (%)	MPP (%)	LGP (%)
Control group	0	0	0	100
A1	10	0	0	90
A2	15	0	0	85
A3	20	0	0	80
A4	25	0	0	75
A5	30	0	0	70
B1	0	10	0	100
B2	0	15	0	90
B3	0	20	0	85
B4	0	25	0	80
B5	0	30	0	75
C1	5	25	0	70
C2	10	20	0	70
C3	15	15	0	70
C4	20	10	0	70
C5	25	5	0	70
D1	20	10	0	70
D2	22	11	0	67
D3	24	12	0	64
D4	26	13	0	61
D5	28	14	0	58
E1	26	13	1	60
E2	26	13	2	59
E3	26	13	3	58
E4	26	13	4	57
E5	26	13	5	56

.

2.3. Sample Preparation

The rigid polyurethane foam was prepared by the chemical foaming method with distilled water and n-pentane as blowing agents. The ratio for preparing RPUF raw material is polyether polyol:MDI:n-pentane:distilled water = 100:100:6:1. The required amounts of distilled water and n-pentane for the weighing test are reserved in measuring cups and recorded as Cup A and Cup B. A certain quantity of MPP, EG, and LGP necessary for the test is placed in Beaker A, and the powdered material is stirred evenly with a stirring rod until the color is uniform and the powder is not clumped. Weigh a specified amount of polyether polyol required for the test into Beaker B, and then pour this material into Measuring Cups A, Measuring Cups B, and Beaker A (now denoted as Beaker C). Use a constant-speed electric mixer to stir the batches of Beaker C at a speed of 500 rpm for 1 min. Weigh the quantity of MDI needed for the test and pour it into Beaker C (now denoted as Beaker D). Use the constant-speed electric mixer to stir the batches of Beaker D at a rotational speed of 1000 rpm for 2 min. Pour the mixture quickly into the prepared mold and allow it to stand for 2 h to fully foam and cure. Since the surface formed by foam curing is not uniform, the sample is trimmed, and the test specimen is prepared. The sample preparation process is illustrated in Figure 2.

2.4. Instrumentation and Measurement

2.4.1. Flame-Retardant Property

The Limiting Oxygen Index (LOI) was utilized to evaluate the flame-retardant performance of the sample. A digital oxygen index tester (JC-107, manufactured by Shanghai Gaozhi Precision Instrument Co., Ltd., Shanghai, China) was employed to measure the LOI during the test, as depicted in Figure 3a. In accordance with the specifications outlined in “Determination of the Burning Behavior of Plastics by Oxygen Index Method—Part 2: Room Temperature Test”, the test sample consisted of fifteen rigid polyurethane foam specimens, each measuring 100 × 10 × 10 mm³. In the process of measuring the LOI, two methods are acceptable for igniting the specimen: the top ignition method and the diffusion ignition method. As per the national standard requirements, since RPUF (Rigid Polyurethane Foam) is a foam material, the top ignition method was adopted. After the specimen was ignited, its combustion length and combustion time were recorded, as presented in

Table 3. LOI record sheet (The table uses A5 as an example).

Initial Oxygen Concentration Determination
Oxygen concentration/%				24.2	25.2		26.2		27.2
Burning time/s				6.4	9.4		12.6		16.8
Burning length/mm				3	4		4		100
reaction (○ or ×)				○	○		○		×
○ indicates burning time ≤ 180 s (or burning length ≤ 50 mm) × indicates burning time > 180 s (or burning length > 50 mm) (the same below)
NT Series Measurement
NL Series Measurement					—				Cf
Oxygen concentration/%	27.2	27.0	26.8	26.6	26.6	26.8	26.6	26.8	26.6
Burning time/s	16.4	17.0	17.3	19.6	19.2	18.1	19.4	20.2	23.3
Burning length/mm	100	100	100	6	8	100	100	20	100
reaction (○ or ×)	×	×	×	○	○	×	×	○	×

.

The initial oxygen concentration test operates as follows: If the combustion behavior of the preceding sample exhibits a “×” reaction, the oxygen concentration is reduced. Conversely, if the combustion behavior of the previous sample displays a “○” reaction, the oxygen concentration is increased. The initial adjustment step in this test is 1%. The test continues until the difference in oxygen concentration falls below 1.0%, and the reactions from two consecutive samples are different, with one being “○” and the other “×”. The oxygen concentration at which the reaction behavior is “○” is defined as the initial oxygen concentration.

N_L series measurement: At this stage, the step size is 0.2% and a specimen is tested using the initial oxygen concentration as the first value of N_L and the reaction behavior is recorded. Subsequent specimens are tested with varying oxygen concentrations, increasing or decreasing in 0.2% increments, until the reaction behavior differs from the first observed reaction. Four samples are tested using this 0.2% oxygen concentration step, and the oxygen concentration of the final sample is denoted as cf. The reaction behavior of the last five tests is documented in

Table 4. The oxygen concentration k value corresponds to the table.

The Reaction of the Last Five Measurements	The k Value of the Initial Measurements, Contingent upon the Specific Reaction Behavior Observed, Serves as an Indicator for the Subsequent Analysis.
	○	○○	○○○	○○○○
×○○○○	−0.55	−0.55	−0.55	−0.55	○××××
×○○○×	−1.25	−1.25	−1.25	−1.25	○×××○
×○○×○	0.37	0.38	0.38	0.38	○××○×
×○○××	−0.17	−0.14	−0.14	−0.14	○××○○
×○×○○	0.02	0.04	0.04	0.04	○×○××
×○×○×	−0.50	−0.46	−0.45	−0.45	○×○×○
×○××○	1.17	1.24	1.25	1.25	○×○○×
×○×××	0.61	0.73	0.76	0.76	○×○○○
××○○○	−0.30	−0.27	−0.26	−0.26	○○×××
××○○×	−0.83	−0.76	−0.75	−0.75	○○××○
××○×○	0.83	0.94	0.95	0.95	○○×○×
××○××	0.30	0.46	0.50	0.50	○○×○○
×××○○	0.50	0.65	0.68	0.68	○○○××
×××○×	−0.04	0.19	0.24	0.25	○○○×○
××××○	1.60	1.92	2.00	2.01	○○○○×
×××××	0.89	1.33	1.47	1.50	○○○○○
	The k value of the initial measurements, contingent upon the specific reaction behavior observed, serves as an indicator for the subsequent analysis.				The reaction of the last five measurements
	×	××	×××	××××
	Corresponding to the reaction in column 6, the value of k given in the table above, but the sign is reversed, the formula is: OI = cf − kd

, and the k value for this group of tests is subsequently obtained.

The oxygen index OI is expressed as a volume fraction and is calculated by Formula (1):

OI = c_f + kd

(1)

OI—the oxygen index

C_f—the oxygen concentration value of the last specimen, expressed as a volume fraction (%), take one decimal place;

k—coefficients obtained from Table 3;

d = 0.2

After oxygen concentration is obtained by Formula (1), in order to reduce the error, the standard deviation of oxygen concentration measurement is obtained by Formula (2).

$$\widehat{\mathsf{\sigma }}={\left[{\displaystyle \frac{\sum _{\mathrm{i}=1}^{\mathrm{n}}{\left({\mathrm{c}}_{\mathrm{i}}-\mathrm{O}\mathrm{I}\right)}^2}{\mathrm{n}-1}}\right]}^{1/2}$$

(2)

$\widehat{\mathsf{\sigma }}$—Standard deviation of oxygen index measurement as calculated by formula;

C_i—N_T series and oxygen concentrations used in each of the two last six reactions;

OI—oxygen index value calculated by Formula (1);

n—The number of measurements

If the standard deviation of oxygen concentration measurement obtained meets Formula (3), the oxygen index is the limiting oxygen index of the sample; otherwise, the error is too large, and it needs to be tested again.

$${\displaystyle \frac{2\widehat{\mathsf{\sigma }}}{3}}<\mathrm{d}<1.5\widehat{\mathsf{\sigma }}$$

(3)

$\widehat{\sigma }$—Standard deviation of oxygen index measurement as calculated by Formula (2).

d = 0.2

2.4.2. Thermal Insulation and Mechanical Properties

Thermal insulation performance test: Thermal conductivity serves as a metric to characterize the thermal insulation performance. For this test, a double-plate thermal conductivity tester (model: IMDRY3001, manufactured by IMBEL (Tianjin, China) Measurement and Control Equipment Co., Ltd., Tianjin, China) was utilized to measure the thermal conductivity, as illustrated in Figure 4a. In accordance with the specifications outlined in “Determination of Steady-State Thermal Resistance and Related Properties of Insulating Materials by the Guarded Hot Plate Method”, the test samples consisted of two rigid polyurethane foam specimens, each measuring 300 × 300 × 30 mm³. To minimize errors, each sample was subjected to testing three times, and the average of these three test results was adopted as the final outcome.

Mechanical properties test: Compressive strength serves as an indicator of mechanical properties. The compressive strength was measured using a microcomputer-controlled electronic universal testing machine (model: WDW-20, manufactured by Shenyang Ziwei Electromechanical Equipment Co., Ltd., Shenyang, China), as depicted in Figure 4b. In compliance with the standards outlined in “Determination of Compressive Properties of Rigid Foam Plastics”, the test samples comprised five rigid polyurethane foam specimens, each with dimensions of 100 × 100 × 30 mm³. The compression rate was set at 3 mm/min, and the test was terminated upon reaching a deformation of 10%. The average of the test results from the five samples was calculated and taken as the compressive strength of the material.

3. Results and Discussion

3.1. Performance Analysis of EG and MPP

Before embarking on the combination study, it is essential to conduct separate investigations into the performance of EG and MPP. These individual studies offer valuable insights into the flame-retardant properties of these additives when used against RPUF, encompassing their flame-retardant efficiency, the impact they have on the physical properties of the material, as well as their compatibility with the substrate. By conducting a comprehensive analysis of the experimental results, we can gain a clear understanding of the advantages and limitations of each flame retardant, which in turn serves as a valuable data reference for subsequent ratio tests.

3.1.1. Performance Analysis of EG

Firstly, the sample was prepared according to the ratio of Part A in Table 1. LOI, thermal conductivity, and compressive strength of the sample were measured. The measurement results are shown in

Table 5. EG performance test results.

Number	LOI (%)	Thermal Conductivity (W/(m·K))	Compressive Strength (kPa)
Control group	18.5	0.030	294.70
A1	21.0	0.046	312.43
A2	22.0	0.054	334.60
A3	22.8	0.048	247.56
A4	24.2	0.057	238.30
A5	26.6	0.050	156.21

, and the variation rule is shown in Figure 5.

LOI variation in EG is shown in Figure 5a. Compared with the control group, the flame retardancy of A1–A5 was significantly improved after the addition of EG. When the addition of EG was 30%, the maximum LOI reached 26.6%, and with the increase in the addition of EG, LOI continued to increase and the rise rate was faster.

The variation in thermal conductivity with respect to EG content is shown in Figure 5b. At an EG content of 10%, optimal thermal insulation performance is reached, with the lowest thermal conductivity recorded at 0.046 W/(m·K). Conversely, when the EG content reaches 25%, thermal insulation performance deteriorates significantly, exhibiting a maximum thermal conductivity of 0.057 W/(m·K). It can be inferred that the incorporation of EG during the preparation of RPUF leads to an increase in thermal conductivity, indicative of a decline in thermal insulation property. Furthermore, as the EG addition increases, the thermal conductivity undergoes a fluctuating pattern. This trend arises due to EG’s acceleration of the foaming reaction, resulting in a higher bubble count and subsequent deterioration of thermal conductivity. However, when EG is added in amounts ranging from 15% to 20%, it fills microscopic voids within the foam, mitigating air convection and thereby reducing thermal resistance. As air is a poor heat conductor, this reduction in convection enhances thermal insulation, lowering thermal conductivity. Conversely, the EG concentrations between 20% and 25% may disrupt the microstructural pore arrangement of RPUF, compromising pore uniformity and integrity. This structural alteration impedes efficient heat transfer, manifesting as poor thermal conductivity on a macroscopic scale. Additionally, it may introduce obstacles or barriers within the thermal conductivity pathways, elevating thermal resistance and further diminishing overall thermal conductivity. Ultimately, as the quantity of expandable graphite increases, it gradually establishes a thermal conductivity network within the RPUF. This network facilitates more efficient heat transfer, shortening the heat transfer path, accelerating the transfer rate, and paradoxically, leading to a reduction in measured thermal conductivity due to the shortened distance over which heat must traverse.

The variation in compression strength of EG is shown in Figure 5c. When the amount of EG is 15%, the compressive strength is the largest, 334.6 kPa, and, when the amount of EG is 30%, the compressive strength is the smallest, 156.21 kPa. As can be seen from the figure, with the increase in EG addition, the compressive strength decreases and then increases, and the decreasing process is lower than that of the control group, with the maximum value reaching 334.60 kPa. However, when EG is 30%, the compressive strength is 156.21 kPa, and the RPUF as a filler material is required by our national standard to be ≥150 kPa. According to the trend of test results, in order to meet the requirements of use, the addition amount of EG should not be greater than 30%.

3.1.2. Performance Analysis of MPP

The RPUF was prepared according to Part B of Table 1, and the LOI, thermal conductivity, and compressive strength of the sample were measured. The measurement results are shown in

Table 6. MPP performance test results.

Number	LOI (%)	Thermal Conductivity (W/(m·K))	Compressive Strength (kPa)
Control group	18.5	0.030	294.70
B1	21.1	0.079	309.35
B2	22.1	0.062	267.20
B3	23.0	0.067	272.45
B4	23.8	0.074	310.82
B5	24.2	0.082	375.43

, and the variation rule is shown in Figure 6.

LOI variation in MPP is shown in Figure 6a. Compared with the control group, the flame retardancy of B1-B5 was significantly improved. When the MPP supplemental level was 30%, the maximum LOI reached 24.2%. With the increase in MPP supplemental level, LOI continued to increase, and the rise rate was slow.

The variation in thermal conductivity of MPP is shown in Figure 6b. With the increase in MPP, the thermal conductivity decreases first and then increases. When the MPP content is 15%, the minimum thermal conductivity reaches 0.062 W/(m·K), and when the MPP content is 30%, the maximum thermal conductivity reaches 0.082 W/(m·K). Compared with the control group, the thermal conductivity increased by 135.6% on average, and the thermal insulation performance deteriorated. When the MPP addition amount is 15–30%, the thermal conductivity increases with the increase in the addition amount, because when the MPP addition amount is too large, it may form obstacles or fractures in the thermal conductivity path inside the RPUF, thereby increasing the thermal resistance, indirectly affecting the thermal conductivity, resulting in poor thermal insulation performance.

The variation in compressive strength of MPP is shown in Figure 6c. With the increase in MPP supplemental level, the compressive strength first decreases and then increases. When the MPP supplemental level is 15%, the minimum compressive strength reaches 267.2 kPa; when the MPP supplemental level is 30%, the maximum compressive strength reaches 375.43 kPa; when the MPP supplemental level is 15% and 20%, the compressive strength reaches 375.43 kPa. The compressive strength was lower than that of the control group. The reason for this is that when the amount of MPP added is low, its dispersion in the RPUF may not be uniform enough. This can lead to excessive MPP concentrations in local areas that may form hard spots or defects during curing, reducing the overall compressive strength of the RPUF. With the increase in MPP content, the dispersion of MPP in RPUF gradually becomes uniform, and the interaction between MPP molecules and the RPUF matrix is enhanced. This helps to form a more dense and stable network structure, which improves the compressive strength of the RPUF. Therefore, the compressive strength decreases first and then increases.

3.2. EG and MPP Matching Experiment

The ratio test of EG and MPP was carried out. The RPUF was prepared according to Part C of Table 1, and the LOI of the sample was measured. The measurement results are shown in

Table 7. MPP-EG performance test results.

Number	LOI (%)
C1	24.4
C2	24.6
C3	26.2
C4	28.2
C5	26.1

.

According to Table 6, C4 is the maximum value in Part C, that is, when EG:MPP = 2:1, the flame-retardant effect is the best, which can reach 28.2%. Part C is compared to Part A and Part B, as shown in Figure 7.

As seen in Figure 7, when the incorporated amount of EG is 5% and MPP stands at 25%, the LOI attains its minimum value of 24.2%, mirroring the maximum value observed in Part B. Conversely, when EG is increased to 20% and MPP reduced to 10%, the LOI achieves its peak at 28.2%, surpassing the maximum of 26.6% recorded in Part A. Hence, blending EG and MPP results in a superior flame retardancy for RPUF compared to when these materials are utilized individually. The underlying mechanism lies in the complementary roles played by EG and MPP: MPP decomposes to emit non-flammable gases and form a coke layer, thereby diluting combustible gases and blocking air, while EG constructs an expanded carbon layer that enhances thermal and oxygen insulation. This synergy refines the flame-retardant system, markedly enhancing its efficacy [26]. The standalone flame retardancy of EG falls short of the synergistic effect obtained with MPP. This is attributed to the inadequate bonding within the expanded carbon layer, preventing the formation of a dense, solid carbon layer that adheres to the material’s surface. Instead, it may “explode” like popcorn upon burning, generating lightweight “fly ash” that can disperse in the air, posing respiratory hazards and compromising the material’s flame retardancy. Incorporating MPP not only preserves EG’s carbonization properties but also emits non-flammable gases to reduce oxygen concentration during combustion, thereby bolstering flame retardancy. Furthermore, the viscous polyphosphate byproducts of MPP reinforce the fluffy carbon layer generated by EG, rendering it denser and more robust [35].

To sum up, it can be seen from the test that the optimal ratio of EG and MPP synergistic action is 2:1. The subsequent tests keep the fixed ratio of EG to MPP, change the overall ratio of EG-MPP, and measure its LOI, thermal conductivity, and compression strength. The test results are listed in

Table 8. EG-MPP performance test results.

Number	LOI (%)	Thermal Conductivity (W/(m·K))	Compressive Strength (kPa)
D1	27.2	0.0498	261.15
D2	28.5	0.0634	210.40
D3	30.0	0.0654	211.26
D4	31.2	0.0612	182.16
D5	--	--	--

.

During the experimental process, where EG and MPP comprised a total of 42% of the mixture, it became challenging to reach uniform stirring within the prescribed time frame due to the increased powder content, as evident in Figure 8. This led to a significant experimental error, and ultimately, the actual amount of flame retardant added fell short of the intended 42%. To uphold the scientific integrity of the test results, the scenario where EG and MPP accounted for 42% was consequently excluded from consideration.

As can be seen from Table 8, when the EG-MPP supplemental amount is 39%, LOI is 31.2%, thermal conductivity is 0.0612 W/(m·K), and compressive strength is 182.16 kPa. In addition to LOI not reaching the experimental target, the thermal conductivity and compressive strength were in line with the national standard, so the addition amount of EG was 26% and the addition amount of MPP was fixed at 13% in the subsequent test.

3.3. EG-MPP-LGP Ratio Test

The ratio test of EG-MPP-LGP was carried out. RPUF is prepared according to Part E in Table 2, as shown in Figure 9.

LOI, thermal conductivity, and compressive strength of the sample were determined, and the test results were listed in

Table 9. EG-MPP-LGP performance test results.

Number	LOI (%)	Thermal Conductivity (W/(m·K))	Compressive Strength (kPa)
D4	31.2	0.0612	182.16
E1	29.5	0.0556	242.18
E2	32.2	0.0555	216.72
E3	32.5	0.0577	183.10
E4	32.4	0.0576	157.34
E5	32.5	0.0603	125.02

.

The changes in LOI, thermal conductivity, and compressive strength of RPUF prepared by mixing three materials are shown in Figure 10.

As can be seen from the chart, when the three flame retardants are mixed to prepare RPUF, when the LGP addition is 3% or 5%, the LOI is the largest, reaching 32.5%; when the LGP addition is 1%, the LOI is the smallest, reaching 29.5%, which is 5% lower than the LOI of D4 group, and the flame retardancy becomes worse. With the increase in LGP addition, LOI increased significantly and stabilized at about 32.4%, and the flame-retardant grade reached B1 level.

When the LGP dosage was 2%, the thermal conductivity was the lowest, 0.0555 W/(m·K), which was 9% lower than that of the D4 group. However, with the increase in the addition amount, the thermal conductivity increases. When the addition amount is 5%, it is 0.0603 W/(m·K), which is the maximum thermal conductivity of the experimental group, but still lower than the thermal conductivity of the D4 group. Note After adding LGP, the thermal conductivity of RPUF decreases overall, and the thermal insulation performance improves. The mechanism is as follows: LGP, as a filler, can fill the tiny pores in RPUF to reduce the path of air convection and heat conduction, and the thermal conductivity of LGP itself is low, which is conducive to reducing the overall thermal conductivity of RPUF.

When the LGP dosage was 1%, the compressive strength was the highest, 242.18 kPa, which increased by 33% compared with the D4 group. The mechanism is that LGP partially melts during the foaming process and penetrates into the cellular structure of the polyurethane foam to form a local strengthening phase. These reinforcement phases can be used as reinforcement points of the cell pillar to improve the overall stability and bearing capacity of the cell structure. In addition, LGP can fill the tiny pores and defects in the polyurethane foam, reduce the weak links in the foam, and form a filling effect, which helps to improve the overall compactness and compressive strength of the foam. With the increase in LGP addition, the compressive strength decreases. When the LGP addition is 5%, the compressive strength is the smallest, which is 125.02 kPa, less than 150 kPa, which does not meet the national standard. Therefore, the E5 composition ratio is not considered in this study.

In summary, according to the comprehensive consideration of flame retardancy, thermal conductivity, and mechanical properties, when the LGP content is 2%, RPUF has the best comprehensive performance, LOI is 32.2%, thermal conductivity is 0.0555 W/(m·K), and compression strength is 216.72 kPa. The results of this study can provide comprehensive properties of RPUF materials. This innovative preparation method provides material support for tunnel safety engineering and has practical value.

4. Conclusions

In the process of in-depth discussion and optimization of flame-retardant properties of RPUF materials, this study is committed to finding an innovative ratio scheme that can not only effectively improve the fire rating of the material, but also maintain or even enhance its original mechanical properties and thermal insulation properties. This study has made progress on the comprehensive analysis of the synergistic mechanism of EG, MPP, and LGP. In the following, the specific results and significance of this matching scheme will be elaborated.

(1) Based on the above detailed ratio test and analysis, it can be seen that the overall ratio of the EG-MPP-LGP intumorable flame-retardant system is 41% of RPUF, and EG:MPP:LGP = 26:13: The thermal conductivity is 0.0555 W/(m·K), the compression strength is 216.72 kPa, and the limiting oxygen index is 32.2%. Compared with pure RPUF, the limiting oxygen index is increased by 74%, and the flame-retardant grade reaches B1 level. Other properties are in line with national standards, and the scheme can be put into practical use.

(2) The expansion of MPP, the heat absorption and oxygen insulation properties of EG, and the melting protection and thermal conductivity characteristics of LGP are superimposed on each other, forming multiple flame-retardant mechanisms that significantly enhance the flame-retardant performance of RPUF. MPP and LGP contribute to the formation of a more continuous, dense, and stable carbon layer structure, while the expansion of EG further increases the thickness density of the carbon layer, thereby more effectively isolating oxygen and heat.

(3) The synergistic mechanism among EG, MPP, and LGP is fully verified in this ratio. They jointly build an efficient and stable expansion flame-retardant system, and this synergistic effect is far more than a single flame-retardant can achieve; it not only improves the flame-retardant performance, but also may in the flame-retardant process through heat absorption and cooling, isolation of oxygen, and the formation of carbon layer mechanisms to provide effective control and suppression of the fire.

(4) The RPUF formulated as a tunnel filling material reduces both the possibility of fire during the construction of the tunnel and the severity of fire after the tunnel is completed. Therefore, the matching scheme has high practical value and promotion prospects, and can be widely used in building insulation, transportation, and other fields, making important contributions to improving public safety.