Influence of Flexibilizers on the Thermal and Combustion Properties of Soundproof Enclosures in Ultrahigh Voltage Converter Transformer Equipment

Abstract

Soundproof enclosures are essential components in ultra-high voltage converter transformer equipment. However, conventional designs pose considerable fire risks, as they may impede fire suppression efforts in case of equipment failure. This study adopted a multi-technique experimental strategy to systematically evaluate the influence of flexibilizer content on the thermal and combustion properties of soundproof enclosures. The methodology combined scanning electron microscopy and thermogravimetric analysis, cone calorimetry and thermal deformation tests. Subsequently, the entropy method was applied to quantify comprehensive fire risk based on the experimental data. The results showed that incorporation of a flexibilizer reduced thermal stability, evidenced by a decrease in the initial pyrolysis temperature from 570 K to 505–545 K at a heating rate of 5 K/min. As flexibilizer content increased, the activation energy (E_α) exhibited a pattern of initial decrease, followed by an increase, and then a subsequent decrease, with most samples exhibiting Eα values below 250 kJ/mol. Simultaneously, flexibilizer addition improved critical fire safety parameters, including reduced heat release rate, total heat release, smoke production, CO₂ release rate, mass loss rate, thermal deformation temperatures, and increased CO release rate. The comprehensive fire risk score decreased significantly from 0.2801 to a range of 0.1147–0.2522 after the addition of the flexibilizer. Thus, this study provides a quantitative assessment of fire safety in ultra-high voltage converter transformer equipment through risk evaluation, offering valuable insights for developing safer enclosure materials.

1. Introduction

To comply with the regulatory requirements set by China’s Ministry of Ecology and Environment, the State Grid Corporation of China has been actively promoting noise control measures in converter stations to mitigate excessive noise levels. Among these initiatives, soundproof enclosure technology has demonstrated high effectiveness in mitigating noise from ultra-high voltage (UHV) converter transformer equipment, thereby helping converter and substations meet noise emission standards [1]. However, the application of such enclosures has also raised significant safety issues. In the event of a fire involving UHV converter transformer equipment, conventional soundproof enclosures can hinder the effective application of fire-extinguishing agents to the source of the fire. This problem is especially acute with top-mounted enclosures, which substantially reduce firefighting efficiency. Such obstructions increase the likelihood of fire spreading to neighboring converter transformer equipment, potentially causing major equipment damage and operational failures [2].

In recent years, design flaws in soundproof enclosures have frequently led to fire incidents at UHV converter stations across China, resulting in substantial economic losses [3]. Notable cases include the following: On 18 June 2018, at the ±800 kV Fulong Converter Station, a failure to promptly open the soundproof enclosure allowed a fire to propagate, damaging three units and resulting in direct economic losses of RMB 240 million. On 5 April 2023, an explosion at the ±800 kV converter transformer in the Ximeng Converter Station generated a shockwave that overturned the enclosure, leading to oil leakage into the valve hall and causing extensive damage to both the roof and the valve tower. These recurring incidents underscore critical safety vulnerabilities inherent in conventional soundproof enclosure designs under fire conditions. Despite the evident risks, research on the fire safety performance of enclosure materials remains inadequate. There is an urgent need to evaluate the thermal stability and fire hazards associated with these structures and to optimize their safety designs. Such work is essential to improving fire protection for converter transformer equipment in UHV systems and will provide vital foundational support for the development of new detachable soundproof enclosures suited to UHV applications.

Scholars worldwide have systematically investigated the thermal stability and fire behavior of various materials, establishing a critical theoretical foundation for elucidating their combustion mechanisms and guiding the development of advanced fire prevention technologies. Tang et al. [4] demonstrated that analyses of pyrolysis and oxidation kinetics can not only assess polymer flammability but also simulate flame propagation behavior, thereby providing valuable insights for the design of high-performance refractory polymers. Onsri et al. [5] employed thermogravimetric analysis to accurately characterize the thermal behavior of fuel samples, revealing correlations between chemical structure and thermal stability. Ding et al. [6] emphasized that pyrolysis serves as a key driver of combustion, noting that the resulting flammable gases sustain gas-phase reactions, and that accurate modeling of this process is essential for predicting fire progression. Fan & Naughton [7] developed a fire development model which showed that combustible gases released from pyrolysis, coupled with sufficient heat feedback, can establish a self-sustaining combustion cycle that accelerates both material degradation and fire hazard. Zhang et al. [8] further elucidated the mechanisms of polymer pyrolysis, which produces both gaseous fuels and solid char under high-temperature conditions. The gaseous products promote flame spread, while the heat from combustion intensifies the pyrolysis process, forming a positive feedback loop [9]. Despite these advances, there remains a notable lack of research focused specifically on the thermal stability and fire behavior of soundproof enclosure materials through pyrolysis-based approaches.

Pyrolysis gases serve as fuel that sustains combustion. Although microscopic pyrolysis characteristics are important, macroscale combustion properties more directly reflect the burning behavior of materials. Among various testing methods, cone calorimetry has become a key technique for evaluating the combustion performance of polymer materials [10]. Zhang et al. [11] verified its broad applicability in analyzing polymer combustion behavior, and Jiang’s research team [12] highlighted its status as one of the most reliable methods for full-scale fire simulations. Lai et al. [13] employed this technology to systematically investigate the combustion characteristics of flame-retardant polypropylene. In a review of its applications, Zhao et al. [14] highlighted that cone calorimetry serves as an effective tool for evaluating material fire resistance, with test data providing critical parameters for assessing flame retardancy in real fire scenarios. Feng et al. [15] specifically pointed out its unique advantage in accurately predicting the ignition behavior of materials under actual fire conditions. Further supporting this, Wang et al. [16] demonstrated its effectiveness in predicting fire intensity in practical situations. Jian et al. [17] endorsed this method as the gold standard in fire testing, providing essential technical support for quantitative flammability analysis by fire safety engineers and researchers. Soundproof enclosures are often employed in environments with fire exposure and pose significant fire risks. However, their combustion properties have not been thoroughly investigated.

The evaluation of material fire safety performance typically requires multidimensional parameter analysis. However, inherent differences in physical meanings and dimensional units among parameters complicate direct comparative assessment. The entropy method effectively addresses this issue by offering an objective approach to standardize multi-source experimental data, thereby enabling scientifically rigorous and quantitative fire risk evaluation. This methodology has been widely adopted in comparative studies of material properties. For instance, Zhang et al. [18] effectively applied the entropy method to evaluate fire hazards in aged insulation materials, obtaining fire risk scores for materials at different aging stages. Similarly, Ma et al. [19] employed this approach to systematically rank fire risks of materials with similar performance, providing data-driven support for engineering material selection. Thus, the entropy method is suitable for analyzing the fire risk of materials involving multiple types and parameters.

Current research on the thermal stability and combustion behavior of soundproof enclosures remains relatively limited. Moreover, existing studies on combustible materials often rely on single-scale experimental approaches, where pyrolysis and combustion characteristics are investigated separately. Given that pyrolysis supplies the fuel for combustion, these processes are intrinsically linked, highlighting the necessity of integrating experimental methods to achieve a comprehensive analysis of material combustion behavior. Furthermore, there has been scarce application of the entropy method for systematic characterization of combustion parameters in fire hazard assessment.

To address these research gaps, this study systematically investigates the thermal stability, combustion performance, and thermal deformation properties of soundproof enclosures using multi-scale techniques, including thermogravimetric analysis, cone calorimetry, and thermal deformation test. Based on these experimental results, the entropy method is employed to integrate multi-parameter data, yielding a quantitative fire risk score that enables an objective and comprehensive evaluation of the material’s fire hazard.

2. Materials and Methods

2.1. Materials

The soundproof enclosure is manufactured by mixing 40–60 mesh gravel with a binder, and a flexibilizer. It is primarily employed in ultra-high voltage converter transformer equipment for noise reduction. The sample preparation process consists of four steps: (1) After weighing the gravel, an appropriate amount of flexibilizer and gelling agent was measured based on the total weight of the gravel. Then, these mixtures were poured into a mixing container and stirred thoroughly until a uniform mixture was achieved. (2) The homogeneous mixture was poured into the pre-weighed gravel and further mixed for 5 min using a mechanical mixer to form a consistent slurry. (3) The resulting slurry was transferred into a sheet-forming mold, compacted and leveled, followed by curing in an oven at 393 K for 45 min. (4) After baking, the sample was allowed to cool naturally to room temperature within the mold before demolding to obtain the final sheet product.

The flexibilizer consisted of dioctyl phthalate, and the binder was composed of polyester polyol. The specific material compositions were detailed in

Table 1. The specific material composition.

Number	Components
1	96.5% gravel + 3.5% binder
2	95% gravel + 3.5% binder + 1.5% flexibilizer
3	94% gravel + 3.5% binder + 2.5% flexibilizer
4	93% gravel + 3.5% binder + 3.5% flexibilizer
5	92% gravel + 3.5% binder + 4.5% flexibilizer

. Sample No. 1, containing only gravel and binder, served as the control group. Flexibilizer was introduced in the other samples (No. 2–No. 5) with contents ranging from 1.5% to 4.5%. The properties of all five samples were compared to evaluate the effect of flexibilizer content.

2.2. Experiments

2.2.1. Thermogravimetry Measurements

Thermogravimetric analysis (TGA) was conducted to evaluate the pyrolysis characteristics of the soundproof enclosures using an SDT Q600 thermal analyzer(headquartered in New Castle, DE, USA). The schematic diagram of the setup is presented in Figure 1 [11]. Experiments were carried out under a nitrogen atmosphere with a constant purge gas flow rate of 100 mL/min. The temperature was raised from 300 K to 1200 K at four different heating rates (5, 10, 20, and 40 K/min). Each sample had a mass of approximately 10 mg. Prior to experiments, the samples were equilibrated in the furnace until a stable mass was achieved. To ensure reproducibility, all experiments were performed in triplicate under each condition.

2.2.2. Cone Calorimetry Measurements

The samples measured 100 × 100 × 20 mm (length × width × thickness). To minimize heat loss and suppress mass transfer at the boundaries, all surfaces except the top face were insulated with aluminum foil. A cone heater was positioned 25 mm above the sample, and a ceramic fiber blanket was placed underneath for additional insulation. A constant external heat flux of 80 kW/m² was applied during the tests. All experiments were repeated three times to ensure reproducibility. A schematic diagram of the cone calorimeter setup is shown in Figure 2 [18].

2.2.3. Thermal Deformation Experiments

The thermal deformation test is a standardized method for evaluating the thermal resistance of polymer materials by determining the heat deflection temperature (HDT). In this experiment, a constant load (calculated using Equation (1)) was applied to the specimen under controlled conditions to induce an initial deformation. The loaded specimen was then placed in a thermal chamber and heated at a uniform rate. As the temperature increased, the mechanical strength of the material declined, resulting in progressive deformation. The experimental setup was illustrated in Figure 3.

The test was terminated when the deformation reached a predefined threshold, as specified by the Chinese National Standard GB/T 1634.1-2019 [20], and the corresponding temperature was recorded as the HDT. This value served as a critical metric for assessing the dimensional stability of the material under thermal load, offering essential guidance for material selection and processing. The standard specimen dimensions used in this test are 80 mm × 10 mm × 5.5 mm.

$$F={\displaystyle \frac{2\sigma b{h}^2}{3L}}$$

(1)

$$\Delta s={\displaystyle \frac{L^2\Delta \epsilon }{600h}}$$

(2)

F represents the applied load. b denotes the sample width. h stands for the sample thickness. Δs is the standard deflection. Δε indicates the bending strain at 0.2%. L refers to the span length between the support contacts. σ is a constant value (0.45 MPa).

2.2.4. Scanning Electron Microscope Experiments

Morphological examination of the samples was conducted using a Hitachi SU8010 scanning electron microscope(Hitachi High-Technologies Corporation, located in Tokyo, Japan). The images were performed at an accelerating voltage of 15 kV with a magnification of ×150.

2.3. Theoretical Methods

2.3.1. Pyrolysis Kinetics

The reaction kinetics of pyrolysis can be mathematically expressed by Equation (3):

$${\displaystyle \frac{d\alpha }{dT}}={\displaystyle \frac{A}{\beta }}f\left(\alpha \right)\exp \left({\displaystyle \frac{-E_{\alpha }}{RT}}\right),$$

(3)

where dα/dT stands for reaction rate. β denotes the heating rate, while f(α) represents the reaction mechanism function. The key kinetic parameters include the pre-exponential factor (A) and activation energy (E_α). T indicates the absolute temperature, and R is the universal gas constant. The conversion rate (α), defined by Equation (4), is calculated from the mass evolution, where m₀, m_∞, and m_t correspond to the initial, final, and instantaneous sample masses, respectively.

$$\alpha ={\displaystyle \frac{m_0-m_t}{m_0-m_{\infty }}}$$

(4)

2.3.2. Distributed Activation Energy Model Method

The Distributed Activation Energy Model (DAEM) method is a widely used isoconversional approach for determining the E_α and A in pyrolysis reactions [21]. It is particularly effective for complex polymers, providing highly accurate kinetic parameters [22]. To overcome the computational complexity associated with the conventional DAEM, a simplified formulation is often employed, as expressed in Equation (5):

$$\ln {\displaystyle \frac{\beta }{T^2}}=\ln \left({\displaystyle \frac{AR}{E_{\alpha }}}\right)+0.6075-{\displaystyle \frac{E_{\alpha }}{RT}}$$

(5)

The DAEM method employs a linear relationship where the kinetic parameters (E_α and A) can be determined by plotting ln(β/T²) against 1/T. The E_α and A are subsequently derived from the slope and intercept of the resulting line, respectively. In this study, the DAEM method was applied to compute the kinetic parameters of the pyrolysis process.

2.3.3. Entropy Method

Based on information entropy theory, the weight of each indicator is determined by its relative contribution to the evaluation system [18]. Indicators demonstrating greater variability correspond to lower information entropy, which implies that they provide more meaningful information and are therefore assigned a higher weight in the comprehensive evaluation [23]. The computational procedure involves the following key steps:

(1)

Data standardization: The raw data are normalized according to the nature of the indicator using distinct equations for forward and reverse indicators to ensure consistent comparability.

$$Y_{ij}={\displaystyle \frac{X_{ij}-X_{min}}{X_{max}-X_{min}}}$$

(6)

$$Y_{ij}={\displaystyle \frac{X_{max}-X_{ij}}{X_{max}-X_{min}}}$$

(7)

The normalized value Y_ij is calculated from the original indicator value X_ij. X_max and X_min represent the maximum and minimum values of the jth indicator, respectively.

(2)

Non-negative transformation. To avoid zero values in subsequent calculations, an infinitesimal offset (0.00001) is introduced to each Y_ij value. This minor adjustment ensures computational stability without compromising data reliability.

(3)

Proportional calculation. The relative contribution (P_ij) of each indicator value is calculated using Equation (8), which quantifies its relative contribution within the overall evaluation matrix. This step generates a standardized distribution across all assessment criteria.

$$P_{ij}={\displaystyle \frac{Y_{ij}}{{\displaystyle \sum _{i=1}^m{Y}_{ij}}}}$$

(8)

(4)

Entropy quantification. The information entropy (e_ij) for each evaluation indicator is determined through Equation (9). A lower entropy value corresponds to a higher information content, indicating greater relative importance of the indicator within the analytical framework.

$$e_j=-k{\displaystyle \sum _{i=1}^m{P}_{ij}\ln \left(P_{ij}\right)}$$

(9)

(5)

Variation coefficient determination. The variation coefficient (g_j) is computed to quantify the relative variability of each indicator. As a critical metric in entropy-based weighting, it reflects the principle that indicators exhibiting greater dispersion provide higher discriminatory power and are therefore assigned increased weight in the evaluation system. The calculation procedure is as follows:

$$g_j=1-e_j.$$

(10)

(6)

Weight determination. The relative importance weight (w_j) for each evaluation indicator is calculated using Equation (11), where each weight is proportional to the indicator’s variation coefficient.

$$w_j={\displaystyle \frac{g_j}{{\displaystyle \sum _{i=1}^m{g}_j}}}$$

(11)

(7)

Composite index calculation. The composite index (sⱼ) is calculated based on the values of P_ij and w_j, as expressed by

$$s_j={\displaystyle \sum _{i=1}^m{w}_j{P}_{ij}}.$$

(12)

3. Results and Discussion

3.1. Structure-Property Relationship

In this section, SEM is employed to characterize the microstructure features of samples with different compositions. When heated, these samples undergo pyrolysis, releasing substantial amounts of combustible volatiles that contribute to the combustion. The pyrolysis and combustion characteristics are evaluated using TGA and cone calorimetry. The TGA results, including thermogravimetric behaviors and kinetic parameters, are presented in Section 3.3. Data from cone calorimetry measurements, such as heat release rate, total heat release, combustion products, and mass loss (ML), are detailed in Section 3.4, Section 3.5 and Section 3.6. To comply with safety requirements, the acoustic enclosure must rapidly deform under thermal exposure and detach from the equipment once ignition occurs. Accordingly, the thermal deformation performance is specifically investigated and discussed in Section 3.7. Finally, a fire risk assessment system is established based on the experimental findings to evaluate the fire hazards associated with the five samples, as described in Section 3.8.

3.2. Microstructure Analysis

As depicted in Figure 4, the microstructure of the soundproof enclosure exhibits a distinct porous structure, contributing to its acoustic damping properties. The porosity of the samples was quantitatively characterized through digital image analysis based on SEM images. The original SEM images were first preprocessed, such as grayscale conversion, to improve image quality. Subsequently, Otsu’s method was applied to perform image binarization to objectively separate pores from the matrix. Finally, the porosity (φ) was calculated as φ = (number of pore pixels/total number of pixels) × 100%.

To ensure statistical reliability, the porosity value for each sample is presented as the mean ± standard deviation obtained from at least three different fields of view. The computational process is implemented in MATLAB 2024 (Version: number: 24.1.0.2537033 (R2024a)). The calculated porosity values for the five samples are 23.27%, 20.86%, 22.77%, 23.63% and 24.45%, respectively. The low porosity of the sample indicates that the flexibilizer, at a content of 1.5%, can be effectively compatible with the binder and gravel, thereby contributing to enhanced sound insulation performance. The order of sound insulation performance is No.2 > No.3 > No.1 > No.4 > No.5.

3.3. Thermogravimetric Analysis

3.3.1. Thermogravimetric Behaviors

Figure 5 shows the pyrolysis characteristics of soundproof enclosures, which is a multi-step reaction.

As demonstrated by thermogravimetric analysis, the incorporation of the flexibilizer significantly modifies the thermal degradation behavior of the material system. The onset temperature for mass loss decreases with the incorporation of the flexibilizer. Since temperature increases linearly with time under programmed heating conditions, this reduction in onset temperature correspondingly shortens the reaction time. Furthermore, the residual mass displays a non-monotonic dependence on flexibilizer content, initially decreasing before increasing with higher concentrations. The pyrolysis process comprises three distinct stages: The first stage is characterized by an ML, primarily attributed to the pyrolysis of the binder. The second stage shows a sharp increase in the MLR, reaching a maximum degradation rate that involves the decomposition of both the binder and flexibilizer. Finally, the third stage demonstrates relative stability, likely due to the pyrolysis of solid residues at elevated temperatures. These observations collectively suggest that the flexibilizer not only modifies the initial decomposition behaviors but also influences the overall thermal stability and residue formation of the material system.

3.3.2. Kinetic Analysis

The E_α represents the minimum energy barrier that must be overcome for the combustion process to occur, serving as a critical indicator of combustion difficulty [24]. Higher values of E_α correspond to increased difficulty in ignition and sustained burning. The calculated E_α values are depicted in Figure 6.

The incorporation of a flexibilizer reduces the energy barrier required for reaction initiation. On one hand, the initial E_α decreases by 147 kJ/mol, from an original value of 380 kJ/mol. On the other hand, the overall E_α throughout the reaction process remains below the initial level of 250 kJ/mol. This reduction in E_α indicates that the reaction initiates more readily in the modified system, accompanied by a decrease in thermal stability. Among these, sample No. 3 exhibits anomalous behavior, with a relatively high overall E_α. Kinetic parameters indicate that this is primarily due to an increased energy demand in the subsequent reaction stages rather than during the initial decomposition phase.

3.4. Heat Release Rate and Total Heat Release

The Heat release rate (HRR) is a critical parameter for assessing combustion behavior, as it provides direct insight into real-scale fire kinetics [22]. Among these, the peak HRR (PHRR) is especially significant due to its major impact on flame spread characteristics, making it a key focus in fire safety research [25]. As shown in Figure 7, experimental results indicate that the incorporation of a flexibilizer effectively reduces the overall heat release rate of the composite material. As flexibilizer content increases, the first exothermic peak becomes progressively less pronounced, and the entire HRR values gradually decrease. It clearly indicates that the flexibilizer contributes to heat release suppression during combustion. Similarly, the total heat release (THR) shows a consistent decreasing trend.

Based on the two parameters, PHRR and time to ignition (TTI), the fire performance index (FPI) can be defined as FPI = PHRR/TTI [26]. A decrease in TTI or an increase in PHRR results in a higher FPI, indicating an increased fire risk. Similarly, the fire growth index (FGI) is calculated as PHRR divided by the time to PHRR (t-PHRR) [27]. A higher FGI value corresponds to either a greater PHRR or a shorter t-PHRR, both of which reflect an elevated fire hazard. The values of FPI and FGI are summarized in

Table 2. The values of FPI and FGI.

Parameters	1	2	3	4	5
FPI-1	1.247	1.152	0.828	0.650	0.062
FPI-2	1.377	1.239	0.992	0.790	0.619
FGI-1	0.808	0.829	0.619	0.492	0.051
FGI-2	0.169	0.167	0.124	0.101	0.081

.

Due to the presence of two peaks in the HRR curve, both FPI and FGI are calculated for each peak. Although the calculated values are different, the observed trends are consistent. As the flexibilizer content increases, both FPI and FGI values decrease. Therefore, from the perspective of heat release, the fire risk of the soundproof enclosures is reduced.

3.5. Combustion Products

The primary hazard in fire scenarios arises from toxic smoke emissions, which pose a dual threat to human survival [28]. Inhalation of these toxic gases can lead to lethality through either direct poisoning or oxygen deprivation. Furthermore, dense smoke significantly reduces visibility and drastically degrades air quality, creating hazardous conditions that impede firefighting efforts [29]. As shown in Figure 8, the smoke production rate (SPR), CO release rate and CO₂ release rate are presented.

Experimental results demonstrate that the flexibilizer effectively suppresses smoke emission. Without the flexibilizer, the sample displays two prominent smoke release peaks. With the addition of the flexibilizer, smoke production is significantly reduced, while the heat release rate shows a steady increase at 3.5% binder content. Notably, although the flexibilizer enhances smoke suppression, it simultaneously promotes the CO release rate. The untreated sample exhibits only a minor initial CO peak followed by a well-defined second peak. As the flexibilizer content increases, the first CO peak progressively intensifies, while the second peak maintains a consistent magnitude. In contrast, CO₂ emission shows an inverse trend, decreasing with higher flexibilizer content, particularly evidenced by the gradual disappearance of the initial CO₂ release peak. These findings reveal a complex interaction between the flexibilizer’s smoke suppression effect and its role in promoting the generation of toxic gases.

3.6. Mass Loss

Figure 9 illustrates the mass loss rate of the soundproof enclosures. Experimental results reveal an inverse relationship between the content of the flexibilizer and the mass loss rate of samples. With increasing concentrations of the flexibilizer, the material exhibits a progressively slower mass loss during combustion. At lower additive levels (below 3.5%), the mass loss rate curve exhibits a distinct bimodal pattern, suggesting two distinct decomposition stages. However, when the concentration of flexibilizer reaches 3.5%, this bimodal behavior transitions into a single, significantly reduced peak, indicating that pyrolysis is more controlled. This modification in mass loss rate profile suggests that the flexibilizer effectively alters the thermal degradation mechanism of the material.

3.7. Thermal Deformation Analysis

The study demonstrates a clear correlation between flexibilizer concentration and the thermal deformation behavior of the material, as illustrated in Figure 10. With increasing flexibilizer content, the temperature required to achieve standard heat deflection exhibits a pronounced decrease, highlighting the substantial influence of flexibilizer on the heat resistance of the materials. At lower concentrations (below 2.5%), the material maintains excellent thermal stability, with HDT ranging from 448 K to 460 K. This performance suggests that the material largely retains its structural integrity and dimensional stability under relatively high-temperature conditions when the flexibilizer content is minimal.

However, a significant transition occurs once the flexibilizer content exceeds 1.5%. The HDT range drops markedly to 314 K–335 K, representing a reduction of approximately 70% compared to samples with lower additive concentrations. This sharp decline clearly indicates that higher concentrations of flexibilizer substantially impair the material’s resistance to thermal stress, resulting in significantly diminished heat resistance characteristics. These findings provide an important research basis for material formulation, indicating that flexibilizer content should be carefully optimized to meet the thermal performance requirements of specific applications.

3.8. Comprehensive Performance Analysis

In the context of pyrolysis, which is a prerequisite for combustion, the pyrolysis behavior serves as an indicator of the material’s fire hazard. The initial pyrolysis temperature (T_i) reflects the thermal stability of the material under heating. A lower T_i implies that pyrolysis is more likely to initiate. T_i can be obtained using the TG-DTG tangent method, as shown in Figure S1. Pyrolytic gases begin to evaporate and mix with air at relatively low temperatures. Once their concentration reaches the lower flammability limit, ignition occurs. The peak mass loss rate (R_p) reflects the generation rate of pyrolytic gases. A higher R_p indicates a faster supply of fuel from the pyrolysis reaction. The temperature at which this peak mass loss rate occurs (T_p) reflects the ease of reaching the maximum decomposition rate. A lower T_p suggests that rapid gas release occurs more readily. The average mass loss rate (R_v) represents the generation rate of pyrolytic gases throughout the entire pyrolysis process. A higher R_v signifies a more intense pyrolysis reaction.

In terms of combustion, the heat released by materials is highly destructive. Thermal hazards arise primarily from the energy released during burning, which can cause structural damage, spread fire, and endanger surroundings. The PHRR and THR are critical parameters that quantify heat release per unit time and over the entire process, respectively. Higher values of PHRR and THR indicate that the material releases heat more intensely and in greater quantity, thereby increasing fire severity. In addition to thermal effects, combustion also generates smoke and toxic gases, such as SPR, CO, and CO₂. A higher smoke release rate leads to more severe fire hazards.

Furthermore, the HDT is indicative of the potential for the acoustic enclosure to impede fire rescue operations. A sufficiently low HDT enables the material to detach promptly from its installed position upon exposure to heat, thereby avoiding obstruction of firewater or fire extinguishing agent spray. Based on these considerations, a comprehensive evaluation system incorporating 10 critical fire performance parameters is established. The raw data and corresponding evaluation results are presented in Table S1 and

Table 3. The estimation results.

Indicators	ej	gj	wj
Ti (K)	0.7852	0.2148	0.0930
TP (K)	0.8159	0.1841	0.0797
RP (×1000 K−1)	0.8505	0.1495	0.0647
Rv (×1000 K−1)	0.6930	0.3070	0.1329
HRR (kW/m2)	0.7801	0.2199	0.0952
THR (MJ/m2)	0.7621	0.2379	0.1030
SPR (g/s)	0.7552	0.2448	0.1059
CO release rate (g/s)	0.7698	0.2302	0.0996
CO2 release rate (g/s)	0.7979	0.2021	0.0875
HDT (K)	0.6796	0.3204	0.1387

, respectively. The composite fire risk scores derived from the analysis for the five samples are as follows: 0.2801, 0.1960, 0.2522, 0.1560, and 0.1147.

The control sample (No. 1) receives the highest score of 0.2801, indicating comparatively greater fire risk. The introduction of the flexibilizer is shown to enhance the fire performance of the material, as evidenced by the reduced scores of the modified samples. These results quantitatively confirm that appropriately formulated flexibilizers can significantly reduce fire hazards. The evaluation system established herein offers a robust methodology for assessing fire risk and supports informed decision-making in material development and formulation optimization.

4. Conclusions

This study comprehensively investigated the influence of flexibilizer on soundproof enclosure materials using an integrated approach that included thermogravimetric analysis, cone calorimetry and thermal deformation experiments. A multi-parameter fire risk evaluation system was established based on the entropy method, enabling quantitative assessment of fire risk across different material formulations. The main findings are as follows:

(1)

The incorporation of a flexibilizer was found to reduce the thermal stability of the samples, as evidenced by three key observations. Firstly, the initial mass loss temperatures decreased from a range of 480–570 K to 370–545 K under five heating rates. Secondly, the activation energy exhibited a complex dependence on flexibilizer content, showing an initial decrease, followed by an increase, and then a subsequent decrease. The vast majority of samples displayed activation energy below 250 kJ/mol in the control sample. Finally, the thermal deformation temperature dramatically decreased from 460 K to 314 K.

(2)

The incorporation of the flexibilizer significantly enhanced fire safety performance by effectively suppressing heat release. The first HRR peak was reduced from 73.57 kW/m² to 3.67 kW/m², and the second peak decreased from 81.25 kW/m² to 36.55 kW/m². The THR values were reduced from 42.13 MJ/m² to 13.43 MJ/m². The average SPR increased from 7.73 m²/s to 11.53 m²/s and then decreased to 7.01 m²/s. The CO release rate for the first peak rose from 1.49 × 10⁻⁴ g/s to 1.66 × 10⁻³ g/s, while the second peak decreased from 1.84 × 10⁻³ g/s to 1.56 × 10⁻³ g/s. Similarly, the first peak of CO₂ release rate dropped from 6.15 × 10⁻² g/s to 6.15 × 10⁻³ g/s, and the second peak decreased from 6.25 × 10⁻² g/s to 2.58 × 10⁻² g/s.

(3)

The overall fire performance scores showed a significant reduction from 0.2801 for unmodified samples to a range of 0.1147–0.2522 for formulations containing the flexibilizer, indicating a reduction in fire hazards. Combustion characteristics and fire risk were influenced by the material composition, allowing flexibility in formulation design based on specific fire safety requirements. Among the tested samples, sample No. 5 (92% gravel + 3.5% binder + 4.5% flexibilizer) exhibited the lowest fire hazards. However, it should be noted that as these materials are intended for use in non-laboratory environments, their long-term performance may be affected by aging. Thus, future studies should consider the effects of aging on material properties.