Simulation and Experimental Study on Parameters of High-Frequency Acoustic Waves Affecting Kitchen Oil Fires

Abstract

This study systematically investigates the influence of sound waves on the flame morphology of oil pan fires in commercial kitchen fire scenarios through a combined approach of numerical simulation and experimental research. A two-dimensional numerical model was established using COMSOL Multiphysics to simulate the interaction mechanisms with flames under various sound source configurations, frequencies, and sound pressure levels. An experimental platform was then constructed to validate and refine the findings using flame morphology, center temperature, and combustion duration as metrics. Results confirm that sound waves effectively destabilize flames, with suppression effects exhibiting a nonlinear trend of initial enhancement followed by attenuation as frequency increases. At the optimal frequency, increasing sound pressure level significantly enhances suppression but exhibits saturation characteristics. Bilateral oblique sound sources simultaneously act on both sides of the flame root, synchronously thinning the boundary layer to create uniform suppression. This configuration also compensates for deflection effects at high frequencies in single-field scenarios, yielding higher efficiency. The determined optimal parameter combination is 1800 Hz and 50 dB, with bilateral oblique arrangement preferred.

1. Introduction

With the rapid development of the national economy, residential fires have become increasingly frequent. Among these, the incidence of kitchen fires and explosions has also risen, causing significant casualties and economic losses. Statistics indicate that approximately 55% of fires in the catering industry originate from commercial kitchens, resulting in severe social impacts and property damage [1]. Commercial kitchens inevitably involve frequent use of high-power appliances and various fuels such as soybean oil, gas, and natural gas. Combined with factors like unattended cooking, improper baking methods, appliance malfunctions, and poor flue heat dissipation, these environments are highly prone to oil pan fires [2].

As traditional fire extinguishing agents face phasing out due to environmental concerns, new fire risks are emerging, driving modern fire suppression technology to evolve. With the enhancement of global environmental regulations, the development of green, efficient, and residue-free fire extinguishing technologies has become a major frontier and research achievement in recent years. Environmentally friendly fire extinguishing methods have gradually replaced polluting agents, showing remarkable innovations in mechanism, material, and application scenarios. Hydrofluoroolefin (HFO) clean agents, exemplified by perfluorohexane, have become the mainstream alternative to traditional hydrofluoroalkanes [3], representing a key innovation in eco-friendly fire extinguishing with low GWP and zero ODP values. However, the risk of hydrogen fluoride generation at high temperatures remains a limiting factor for their application, especially in high-temperature oil fires in commercial kitchens. In contrast, inert gas extinguishing agents offer core advantages of leaving no chemical residues post-extinguishment and being environmentally friendly [4], which is a typical achievement of clean fire suppression technology. However, their limitations include large system footprints, complex designs, and higher costs, which restrict their popularization in small spaces such as kitchens. Fine water mist technology emerged as a novel extinguishing technique in the 20th century due to its multiple effects: cooling through water evaporation heat absorption, suffocation by displacing oxygen with water vapor, and attenuation of thermal radiation [5]. As a widely recognized eco-friendly innovation, fine water mist has achieved fruitful progress in both fundamental research and engineering applications. Furthermore, coupling water mist with chemical additives effectively extinguishes electrical fires, demonstrating superior performance [6]. However, its efficiency heavily depends on specific parameters [7], such as droplet size, flow rate, and arrangement, and non-full-coverage protection modes may create blind spots. Aerosol extinguishing technology, as a non-pressurized storage solution, releases ultrafine solid salt particles and gases to suppress combustion chain reactions through intense chemical reactions [3], representing another innovation in compact and flexible fire suppression. However, some hot-melt aerosols generate high temperatures during release, posing potential harm to humans [8]. These limitations restrict their application in highly sensitive environments. Given the above achievements, innovations, and bottlenecks of existing eco-friendly fire extinguishing methods, non-traditional physical fire suppression technologies represent emerging research directions. Low-frequency acoustic fire suppression, as a novel technique that disturbs flames and disrupts stability, with zero pollution, zero residue, and no chemical reactions, provides a new environmentally friendly innovation for special fire scenes and offers new rescue approaches for specialized fire scenarios [9].

2. Acoustic Wave Fire Suppression Technology

The interaction between sound waves and flames constitutes a complex multiphysics process at the intersection of fluid mechanics, thermodynamics, acoustics, and chemical kinetics. This interaction can significantly alter flame structure, stability, and combustion characteristics, potentially leading to flame extinction [10,11,12].

2.1. Sound Pressure Oscillations and Flame Response

As a nonlinear dynamic system with inherent oscillatory properties, flames exhibit diverse responses to external sound pressure perturbations. At the macroscale, sound pressure oscillations induce periodic volumetric changes throughout the flame. For diffusion flames, excessive expansion and contraction driven by sufficiently large sound pressure amplitudes can tear the flame front, disrupting the flame’s continuous structure [13]. At the mesoscale, volume forces generated by sound pressure gradients act on the flame front, driving motion of the flame surface and surrounding gas. This induces periodic bending and wrinkling of the flame front, significantly increasing its surface area. When the heat loss rate due to increased surface area exceeds the chemical reaction’s heat release rate, the flame may become unstable and cease propagation [14]. The degree of flame response to acoustic pressure oscillations depends on the matching relationship between acoustic wave parameters and intrinsic flame characteristics. The dimensionless Strouhal number ($\mathrm{S}\mathrm{t}=\mathrm{f}\mathrm{L}/\mathrm{U}$) is a key parameter; flames exhibit maximum sensitivity to acoustic perturbations when St~O(1) [15]. The changes in flame morphology under the influence of acoustic waves are shown in Figure 1.

2.2. Acoustic Waves and Their Fire Suppression Effects

Acoustic streaming is a steady-state or low-frequency secondary flow phenomenon generated when sound waves propagate through viscous fluids. Its velocity is typically much lower than the particle vibration speed of the sound wave, yet it maintains a constant direction, enabling sustained fluid transport effects [16]. In acoustic flame extinction scenarios, acoustic streams play multiple critical roles: First, they enhance convective heat transfer, increasing the convective heat transfer coefficient by more than an order of magnitude. This rapidly removes heat from the fuel surface, reducing fuel evaporation or pyrolysis rates and fundamentally diminishing the supply of combustible gases [11]. Second, they alter local mixing characteristics. Acoustic flows generate complex vortex structures within the flame region, intensifying turbulent mixing between fuel vapors and air. This may create areas of excessively lean or rich mixtures, forming localized extinguishment points [17]. Third, the blow-off effect occurs when the acoustic flow velocity equals or exceeds the flame propagation speed, physically stripping the flame front from the fuel source or overstretching it to rupture [12]. Fourth is disturbance of the flame base. The flame root is one of the most vulnerable parts of the flame; even minor flow disturbances can cause the flame to transition from a stable attached state to an oscillating or detached state [13].

The action intensity and manifestation of acoustic waves are closely related to the type of flame and fuel, and obvious differences can be observed among fire classes. For gaseous fuel flames, the reaction zone is thin and the flow field is relatively uniform. Acoustic waves can easily cause strong stretching, shear, and blow-off of the flame surface. The flame is highly sensitive to periodic perturbation and tends to be extinguished by acoustic-induced flow distortion. For liquid fuel flames (e.g., kitchen oil pan fires), which are the focus of this study, the suppression mechanism is more complicated. Acoustic waves change the evaporation rate of liquid fuel, the mixing process between fuel vapor and air, and the heat feedback from the flame to the fuel surface. The flame base and thermal boundary layer are easily disturbed, resulting in instability, oscillation, and even extinction. For solid fuel flames, the combustion process is dominated by pyrolysis and char combustion. Acoustic waves mainly enhance heat dissipation and promote local turbulence mixing, thereby weakening the combustion intensity and slowing down the flame spread rate. In general, acoustic waves achieve flame suppression through four typical effects. Enhancing convective heat transfer to cool the fuel surface and reduce fuel volatilization. Intensifying turbulent mixing to form overly lean or overly rich local mixtures that cannot sustain combustion. Producing a hydrodynamic blow-off effect when the acoustic flow velocity exceeds the flame propagation velocity. Directly disturbing the stable flame root to destroy its attachment and cause liftoff or extinction. The following are relevant studies on the effects of sound waves on different types of flames, as shown in

Table 1. Summary of representative studies on acoustic flame extinguishing for different fuel types.

Fuel Type	Representative Authors	Research Object	Core Findings
Gaseous Fuel	Niegodajew et al. [18]	Methane jet flame	Acoustic oscillation with St~O(1) can effectively quench gaseous jet flames by stretching the flame front and enhancing heat loss.
Gaseous Fuel	Xiong et al. [13]	Propane diffusion flame	Low-frequency acoustic waves (50–70 Hz) show better suppression effect on gaseous flames than high-frequency waves.
Gaseous Fuel	Plascencia et al. [19]	Turbulent methane flame	At a pressure anti-node, the coupling of the acoustics and flame gave rise to an axisymmetric response, which prompted the flame to become unstable at the anchoring region.
Liquid Fuel	Xiong et al. [11]	Dripping liquid flame	Acoustic waves can disrupt the thermal boundary layer of liquid fuel, reducing evaporation and inducing flame extinction.
Liquid Fuel	Shi et al. [20]	Pool fire	Tests on an ethanol pool fire perturbed by acoustic waves were conducted at acoustic frequencies of 20–100 Hz and acoustic pressures of up to 1.2585 Pa.
Solid Fuel	Wilk-Jakubowski [21]	Candle containing paraffin wax	Using a high-power acoustic fire extinguisher and an ignition source (specifically, a paraffin candle), the study analyzed how the order of the even-order harmonics affects the acoustic fire extinguishing process.
Solid Fuel	Xiong et al. [12]	Dry wood ball	The cumulative effect of firebrand motion and acoustic oscillation was found to facilitate flame extinction. A characteristic Damköhler number (~1), is used to quantify the extinction limit of the flaming firebrand.
Solid Fuel	Loboichenko et al. [22]	Candle	Low-frequency modulated and unmodulated acoustic waves generated by a 1700 W high-power acoustic extinguisher can effectively extinguish candle flames.

.

2.3. Coupling of Acoustic Resonance and Combustion Instability

When acoustic frequencies match the inherent acoustic or hydrodynamic modes of a combustion system, intense resonance can be triggered, significantly amplifying the disturbance effect of sound waves on the flame [16], as shown in Figure 2.

The acoustic modes of a combustion system are determined by its geometric boundaries. When the frequency of an external sound source approaches these natural frequencies, acoustic resonance occurs within the cavity, significantly amplifying the sound pressure amplitude and greatly enhancing the intensity of the acoustic wave’s effect on the flame. To quantitatively describe the propagation and resonance behavior of acoustic waves in the flow field, the pressure acoustic wave equation is introduced as follows. For a complete list of symbol definitions, see Appendix A.

$$\nabla ·\left[-{\displaystyle \frac{1}{{\rho }_c}}\left(\nabla P_t-q_d\right)\right]-{\displaystyle \frac{k_{cq}^2{p}_t}{{\rho }_c}}=Q_m$$

(1)

More critically, a feedback coupling exists between the heat released during combustion and the sound pressure oscillations, known as thermoacoustic instability [23]. This process can be described as follows: initial acoustic pressure perturbations alter mixing and flow, modulating heat release rates and generating new acoustic pressure oscillations. When heat release pulsations are in phase with pressure pulsations and gain exceeds loss, the system undergoes self-oscillatory behavior. In acoustic flame extinction, active acoustic sources can be employed to enhance this coupling, rapidly driving the flame toward instability until extinction [16].

Under resonant conditions, even a moderate sound source can induce strong flame deformation, local extinction, and overall blow-off, which is especially significant for liquid fuel fires such as kitchen oil pan fires.

2.4. Thermoacoustic Interaction and Combustion Chemistry

Acoustic waves influence flames not only through hydrodynamic mechanisms but also modulate combustion via thermodynamic and chemical pathways [15]. Adiabatic temperature fluctuations induced by sound waves directly alter reaction rates. Resonance effects may emerge when acoustic periods are comparable to characteristic chemical timescales, significantly modifying combustion characteristics. Acoustic waves also affect heat transfer by perturbing thermal boundary layers. Oscillatory flow can thin the thermal boundary layer, significantly enhancing heat loss from fuel surfaces [11]. For liquid fuel fires, acoustic waves directly increase evaporation rates by enhancing convective heat transfer coefficients, yet they simultaneously reduce evaporation rates by diminishing thermal feedback from the flame to fuel surfaces. The net effect on evaporation depends on the competition between these two opposing effects [14]. The competing mechanisms by which sound waves affect liquid fuel evaporation are shown in

Table 2. Competing mechanisms of acoustic waves on liquid fuel evaporation.

Mechanism	Effect on Evaporation	Dominant Condition	Corresponding Acoustic Parameters
Convective heat transfer enhancement	Increase (Promote evaporation)	Low frequency, low sound pressure	Frequency: 500–1000 Hz; Sound Pressure Level (SPL): 30–50 dB
Flame thermal feedback weakening	Decrease (Inhibit evaporation)	High frequency, high sound pressure	Frequency: 1500–2500 Hz; Sound Pressure Level (SPL): 50–70 dB

below.

Combining these mechanisms, when disturbance intensity exceeds the flame’s stable recovery capability, flame extinction occurs. Critical extinction conditions correspond to the threshold where these mechanisms collectively act, typically represented as extinction boundary curves on the frequency–sound pressure level plane [10,12]. As shown in Figure 3. It is reasonable to assume that the various parameters affecting flame stability all follow certain patterns of influence.

2.5. Related Studies on the Effects of Sound Waves on Flames

Numerous existing studies have explored the interaction mechanism between acoustic waves and different types of flames, laying a solid theoretical and experimental foundation for the research of acoustic fire suppression technology. These representative studies, partially summarized in Table 1, have clarified the basic laws of acoustic waves on flame behavior, and their core findings are systematically sorted out as follows:

Rayleigh [23] observed that sound waves can affect flame height, with a significant decrease occurring when the excitation frequency matches the resonance frequency. This phenomenon is consistent with the acoustic resonance mechanism discussed in Section 2.3, where resonance amplifies the disturbance effect of acoustic waves, thereby inducing a more obvious reduction in flame height. Oh et al. [24] experimentally demonstrated that acoustic excitation at resonance frequency promotes coherent vortex structures that enhance air entrainment and fuel–air mixing, while simultaneously inducing periodic fluctuation of the liftoff height. Lee [25] similarly found that when the sound frequency coincides with the combustion system frequency, the flame height reduction is accompanied by increased turbulence at the outlet. This enhanced turbulence further promotes the mixing of fuel vapor and air, intensifies heat loss, and jointly contributes to flame instability. In fact, flame height under acoustic excitation does not always decrease; height reduction occurs only within a specific frequency range. This frequency dependence is closely related to the inherent oscillation characteristics of the flame system and the acoustic parameters, which also explains the different suppression effects of acoustic waves on flames under various working conditions. Steinbacher [26] and colleagues investigated the mechanism of acoustic wave interaction with flames, supplementing the theoretical understanding of their mutual influence. Their research further revealed the coupling relationship between acoustic streaming, heat transfer perturbation, and combustion chemical reactions, providing a more comprehensive theoretical framework for the analysis of acoustic fire suppression mechanisms. Niegodajew et al. [18] experimentally demonstrated that lower acoustic frequencies exhibit more pronounced flame suppression effects, and the acoustic pressure influencing the flame remains constant regardless of distance variation. This finding provides important guidance for the selection of acoustic parameters in practical fire suppression applications, especially for the layout of acoustic sources. Xiong et al. [11] employed comparative methods to observe dripping flames under acoustic influence, discovering that suppression effectiveness varied with flame height and falling velocity. Increasing sound pressure enhanced flame extinguishing. Their research focused on liquid fuel flames, which is highly consistent with the research object of this paper (kitchen oil pan fires), and their findings provide direct experimental support for the subsequent experimental design and parameter optimization of this study. Yılmaz-Atay and Wilk-Jakubowski [27] reviewed environmentally friendly fire extinguishing approaches ranging from intrinsically flame-retardant materials and nanocomposites to acoustic oscillations, discussing measurement results obtained with high-power acoustic extinguishers and highlighting the potential of appropriately timed acoustic waves for flame suppression. Vovchuk et al. [28] reviewed the application of acoustic effects in extinguishing oil and petroleum product fires, highlighting its environmental friendliness and development prospects. Yadav et al. [29] designed a sound wave fire extinguisher and experimentally determined that flames of various types can be extinguished within the frequency range of 40–60 Hz. According to Stawczyk et al. [30], non-invasive flame extinguishment using a high-power acoustic extinguisher was validated across varying distances and acoustic frequencies. These findings collectively demonstrate acoustic suppression of flames under specific conditions, prompting recent proposals for acoustic fire suppression applications. However, it is worth noting that most of the existing studies are carried out under ideal laboratory conditions, which are quite different from actual indoor fire scenarios. Specifically, indoor fires involve complex scenarios unlike the isolated conditions of laboratory experiments, potentially featuring multiple ignition points, diverse flame types, and various heat transfer and combustion media. Therefore, relying solely on acoustic waves for fire suppression requires further investigation into factors such as sound pressure, frequency, and duration affecting flame combustion. Undeniably, acoustic fire suppression technology holds broad application prospects and practical value for indoor fires, especially for kitchen oil pan fires with the characteristics of small scale, high frequency, and easy recurrence. The research gaps existing in current studies (such as the lack of research on high-frequency acoustic waves in complex indoor scenarios) are exactly the core content that this paper aims to supplement and improve, which is of great significance for promoting the engineering application of acoustic fire suppression technology.

3. Numerical Simulation of High-Frequency Acoustic Waves on Grease Fires

COMSOL Multiphysics (Version 6.1, COMSOL Inc., Burlington, MA, USA) features multiphysics simulation technology. In particular, its pressure acoustics module can accurately simulate the complex propagation characteristics of sound waves from low to high frequencies. It can not only present dynamic sound fields but also construct high-precision indoor environments, and it is capable of simulating indoor acoustic environments while coupling with heat transfer and fluid physics.

It is important to acknowledge that a substantial body of literature has demonstrated superior flame extinction performance using low-frequency acoustic waves, typically in the range of several tens of hertz [18,20,24,30]. For instance, Niegodajew et al. [18] showed that lower frequencies (30–50 Hz) enable easier flame extinguishment with less acoustic power, and Shi et al. [20] reported that 20–100 Hz waves induce pronounced suppression of pool fires. These findings are well established and are not contradicted by the present work. Nevertheless, the present study focuses on the high-frequency range (500–3000 Hz) for two practical engineering reasons. First, low-frequency waves have long wavelengths and strong penetrating power, which can easily excite structural resonance in building components and generate excessive low-frequency noise in enclosed indoor environments (e.g., commercial kitchens), making them less suitable for real-world fire suppression applications. Second, high-frequency waves offer better directivity and can be more easily attenuated or shaped to avoid unwanted reflections and resonance in confined spaces, which is advantageous for targeted flame disturbance. A similar engineering trade-off has been discussed in the context of acoustic agglomeration and fire suppression [31]. A limitation of this study is that the chosen high-frequency range is intrinsically less efficient in flame extinction than low-frequency waves. Nonetheless, it offers practical advantages in kitchen environments.

Although laboratory studies have confirmed that low-frequency acoustic waves possess excellent fire-extinguishing properties, Zheng’s research team [31] found that acoustic excitation at 380–500 Hz induces periodic oscillations in laminar flames and may promote flame reignition through Kelvin–Helmholtz vortices. Shi et al. [32] proposed a modified model and revealed a threshold effect of 20–100 Hz sound waves on the fuel consumption rate: low frequencies (<80 Hz) tend to induce flame extinction, while higher frequencies lead to complex reignition dynamics. Therefore, taking into account both the established knowledge of low-frequency efficiency and the practical constraints of kitchen environments, this study employs a frequency range of 500–3000 Hz to analyze fluid temperatures in commercial kitchen fires, simulate flame patterns in cooking oil fires, and investigate the effects of plane wave acoustic fields with varying parameters between two placement methods (top-facing and side-tilted). The impact of the disturbance on flame patterns is summarized through velocity distributions.

3.1. Model Development and Mesh Independence Verification

Since this study investigates the effects of acoustic waves with varying parameters on flame morphology, primarily observing motion trajectories along the longitudinal plane, a 2D model was selected to reduce computational load and minimize errors. The interface was chosen to establish a two-dimensional model as shown in Figure 4. The room is a 10 m wide by 5 m high rectangle, representing a commercial kitchen, designated as Domain 1. A commercially available stove, typically 1.25 m wide and 0.85 m high, is considered. To simulate the actual ignition point, a small rectangle 0.5 m wide and 1 m high is placed at the center bottom of the stove, representing the ignition source (designated as Domain 2). The top surface of this rectangle is defined as the combustion surface and smoke dispersion surface. Together, they form a unified domain (Domain 1). Air (Air) is selected from the built-in material library as the additional material, with default properties applied. Since the simulation involves oil pan fire smoke with temperature variations during flow, higher-precision heat flux modeling is required due to increased fluid velocity at elevated temperatures. Therefore, the Fluid Heat Transfer (ht) and Turbulence, Low Reynolds Number ($\mathrm{k}-\mathsf{\epsilon }$, spf2) modules were selected. These were coupled using the Multiphysics: Non-Isothermal Flow (nitf2) interface. And an external acoustic field was set in the Pressure Acoustics, Frequency Domain (acpr) module. In this simulation, the oil pan fire is simplified to a thermal plume model at low Mach numbers, and a low-Reynolds-number turbulence model (based on laminar flow approximations) is adopted to ensure computational stability and efficiency, with a focus on macroscopic thermo-fluid–acoustic coupling interactions. Although these simplifications deviate somewhat from actual fire scenarios, they effectively reduce computational complexity. From a macroscopic perspective, they allow for the derivation of parameter ranges, providing a reference for subsequent experimental work to identify the optimal operating parameters. The current model does not yet fully account for key mechanisms closely related to pressure acoustics, such as detailed physicochemical reactions and turbulence-induced quenching. In future work, these complex coupling mechanisms will be further refined through the development of high-resolution three-dimensional simulations.

To clearly illustrate the modeling and solution process, we now provide a systematic summary of the simulation steps for each section:

The heat transfer in fluids (ht) interface was applied to domain 1 with an initial temperature of 293.15 K. A random temperature fluctuation function was introduced to mimic the intermittent temperature variation (≈400 °C) during ignition. Open boundaries were set on both lateral sides and an exhaust boundary at the top to represent typical kitchen ventilation.

$$d_z\rho C_{p}u\times \nabla T+\nabla q=d_{z}Q+q_0+d_z{Q}_p+d_z{Q}_v$$

(2)

$$q=-d_{z}k\nabla T$$

(3)

For fluid flow, the laminar flow (spf) interface was used with the RANS-based low Reynolds number k-ε turbulence model (automatic wall treatment). No-slip condition was imposed on solid walls, and gravity was included as a body force. The left and right boundaries were open with non-viscous stress, and the top was a pressure outlet at 0 Pa to prevent backflow. The non-isothermal flow (nitf) coupling combined the heat transfer and laminar flow interfaces over domain 1, incorporating viscous dissipation and buoyancy effects.

$$\rho \left(u2·\nabla \right)u2=\nabla ·\left[-p2I+K\right]+F+\rho g$$

(4)

$$\rho \nabla ·u2=0$$

(5)

$$K=(\mu +{\mu }_T)\left(\nabla u2+{\left(\nabla u2\right)}^T\right)$$

(6)

$$\rho \left(u2·\nabla \right)k2=\nabla ·\left[\left(\mu +{\displaystyle \frac{{\mu }_T}{{\sigma }_k}}\right)\nabla k2\right]+P_k-\rho ϵ$$

(7)

$$\rho \left(u2·\nabla \right)ϵ=\nabla ·\left[\left(\mu +{\displaystyle \frac{{\mu }_T}{{\sigma }_{ϵ}}}\right)\nabla ϵ\right]+c_{ϵ1}{\displaystyle \frac{ϵ}{k2}}{P}_k-c_{ϵ2}\rho {\displaystyle \frac{{ϵ}^2}{k2}}{f}_{ϵ}(p,\mu ,k2,ϵ,l_w), ϵ=ep2$$

(8)

$$\nabla G2·\nabla G2+{\sigma }_{w}G2(\nabla ·\nabla G2)=(1+2{\sigma }_w){G2}^4,l_w={\displaystyle \frac{1}{G2}}-{\displaystyle \frac{l_{ref}}{2}}$$

(9)

$${\mu }_T=\rho C_{\mu }{\displaystyle \frac{{k2}^2}{ϵ}}{f}_{\mu }(\rho ,\mu ,k2,ϵ,l_w)$$

(10)

$$P_k={\mu }_T\left[\nabla u2:\left(\nabla u2+{(\nabla u2)}^T\right)\right]$$

(11)

Using the Non-Isothermal Flow (nift) multiphysics coupling, the Laminar Flow (spf) interface for fluid flow and the Fluid Heat Transfer (ht) interface for heat transfer were configured, both applied to the default domain 1.

$$Q_{vd}=\tau ·\nabla u$$

(12)

$$F_g={\rho }_{ref}\left(1-{\alpha }_{p,ref}\right)\left(T-T_{ref}\right)g$$

(13)

$${\alpha }_p=-{\displaystyle \frac{1}{\rho }}{\left({\displaystyle \frac{\mathsf{\partial }\rho }{\mathsf{\partial }T}}\right)}_p$$

(14)

$$-n·q=\rho C_p{u}_{\tau }{\displaystyle \frac{T_w-T}{T^{+}}}$$

(15)

$$T^{+}=\{\begin{array}{c}Pr y^{+}, if y^{+}<{\displaystyle \frac{10}{{Pr}^{1/3}}}\\ 15{Pr}^{2/3}-{\displaystyle \frac{500}{{\left(y^{+}\right)}^2}}, if {\displaystyle \frac{10}{{Pr}^{1/3}}}\le y^{+}<\sqrt{1000{\displaystyle \frac{\kappa }{{Pr}_T}}}\\ {\displaystyle \frac{1}{\kappa }}{Pr}_T\mathit{ln}\left(y^{+}\right)+15{Pr}^{2/3}-{\displaystyle \frac{1}{2\kappa }}{Pr}_T\left(1+\mathit{ln}\left({\displaystyle \frac{1000\kappa }{{Pr}_T}}\right)\right), if y^{+}\ge \sqrt{1000{\displaystyle \frac{\kappa }{{Pr}_T}}}\end{array}$$

(16)

The acoustic field was introduced via the Pressure Acoustics, Frequency Domain (acpr) interface. Air was used as the medium (293.15 K, 1 atm). A plane wave radiation condition with inward propagation was applied at the upper edge, with sound speed 343 m/s and pressure amplitude sweeping from 0 dB to 100 dB. All other boundaries were set as sound hard walls.

$$\nabla ·\left(-{\displaystyle \frac{1}{{\rho }_c}}\left({\nabla }_{P_t-q_d}\right)\right)-{\displaystyle \frac{k_{eq}^2{p}_t}{{\rho }_c}}=Q_m$$

(17)

$$p_t=p+p_b$$

(18)

$$k_{eq}^2={\left({\displaystyle \frac{\omega }{c_c}}\right)}^2-k_z^2$$

(19)

A physics-controlled mesh with normal element size was used, and all physical fields were included in the mesh generation. The resulting mesh distribution, shown in Figure 5, was refined in regions where flue gas was released and temperature increased, while a uniform distribution was applied elsewhere. For the pressure acoustics setup, domain 1 was selected with no port sweep and normalized size. Air was used as the reference pressure medium. The pressure acoustics interface was configured with a temperature of 293.15 K, an absolute pressure of 1 atm, with density and speed of sound taken from the material properties. A hard acoustic boundary (wall) was applied to all boundaries except the outlet. A plane wave radiation condition was imposed on the upper edge (inward propagation), with the sound velocity set to 343 m/s and the pressure amplitude ranging from 0 dB to 100 dB.

The mesh was created using COMSOL’s built-in meshing functionality. However, the accuracy of numerical simulation results is closely tied to the size of the generated mesh. Within the acceptable error range of results, progressively refining the mesh enhances data precision and realism. Simultaneously, computational complexity increases, leading to higher computational rounding errors. Therefore, the independence of the mesh was evaluated. Simulations of the variable-temperature fluid field at 1500 Hz were performed using three mesh sizes—standard, refined, and ultra-refined—to obtain temperature variations at point probes. The results are shown in Figure 6.

The mesh independence test confirmed that the refined mesh scheme provided stable results and was selected for all simulations. The solution procedure comprised a frequency-domain sweep (500–3000 Hz, 500 Hz steps) followed by a transient analysis with output time steps of 0.05 s over 5 s. The fluid flow variables were solved using the transient study, with unsolved variables taken from the frequency-domain solutions, as shown in Figure 7.

3.2. Effect of Different Sound Source Arrangements on the Flame

By plotting the temperature and flow velocity results, as shown in Figure 8, the evolution of the flame under acoustic influence can be observed. Without an acoustic field, the temperature gradient begins to diffuse laterally between 2 and 2.5 s and continues to develop over the next two seconds. By 4.5 s, lateral diffusion stabilizes, though flame height continues to increase slightly until reaching the state shown at 5 s, as shown in Figure 8a. When the applied sound field frequency reaches 500 Hz, subtle changes emerge in the temperature flow, as shown in Figure 8b, with the top region exhibiting signs of rightward displacement, indicating compromised flame stability. At 1000 Hz, the temperature flow undergoes a noticeable transformation. As depicted in Figure 8c, downward pressure appears at both ends, indicating suppression of the flame’s vertical development. Beyond 1500 Hz, as shown in Figure 8d, the acoustic field’s influence becomes more pronounced, causing the temperature field to elongate. This signifies that the flame’s lateral spread is also being suppressed, thereby preventing further expansion of the fire.

When the applied acoustic wave frequency reaches 2000 Hz, significant turbulence appears in the temperature flow, as shown in Figure 8e. The flame shape deflects to the right, its apex structure is disrupted, and temperature fluctuations increase on both sides, demonstrating the enhanced effect of acoustic waves on flame morphology. When the acoustic frequency increased to 2500 Hz, as shown in Figure 8f, the acoustic force exerted on the plane became very strong, nearly dispersing the main flame body. However, temperature diffusion continued to increase, and the flame morphology did not exhibit fragmentation. Applied to actual combustion scenarios, this should represent an excitation-assisted combustion effect. Compared to 2500 Hz, when the sound wave frequency reached 3000 Hz, the main flame body became more regular, and the flame height changed. Compared to the 1500 Hz effect, both vertical and horizontal directions were stimulated and developed, as shown in Figure 8g.

When the emission surface is set to the left and right, consideration of the conflict of the intersecting boundaries and the inconsistency of the time step that cannot be calculated is needed to ensure that its center plumb line intersects with the smoke generation surface, and the rectangle is changed to a polygon as shown in Figure 9. The rest of the parameter settings are with the above forward sound source.

The temperature–velocity field results as shown in Figure 10 reveal that lateral sound sources exert significantly greater influence on temperature and fluid dynamics compared to frontal sources. As shown in Figure 10a, at 500 Hz, the lateral source nearly matches the effect of a 1000 Hz frontal source, suppressing longitudinal flame propagation at both ends and inducing a downward pressure state. However, as the frequency increases to 1000 Hz, as shown in Figure 10b, the flame control effect is not enhanced but remains comparable to that of the forward-facing source. Even when the frequency further increases to 1500 Hz, as shown in Figure 10c, the flame influence remains unchanged, characterized by longitudinal compression of the flame body and suppression of temperature diffusion at the upper sides. At 2000 Hz, as shown in Figure 10d, the flame primarily deflects leftward, with a reduced flame body height. Although the leftward longitudinal diffusion velocity increases, it still exhibits a suppressing effect.

As the acoustic frequency further increases to 2500 Hz, Figure 10e reveals that the acoustic force acting on the plane becomes stronger than that from a frontal source. The maximum fluid velocity exceeds that observed at 2500 Hz with a frontal source, resulting in a taller flame body. The right side of the flame exhibits a fracture phenomenon, known as splashing. At this frequency, the acoustic wave transitions from a flame-suppressing effect to a promoting effect, which manifests as enhanced combustion intensity rather than extinction. This transition occurs because high-frequency acoustic disturbance intensifies turbulent mixing and increases the entrainment of ambient air into the combustion zone, thereby improving oxygen supply and accelerating combustion reactions. Including 2500 Hz in the analysis is intended to reveal the complete evolution trend of flame behavior from suppression to excitation, so that the critical frequency threshold between these two regimes can be clearly identified. At 3000 Hz, the flame body deflects to the right. When applied to actual combustion phenomena, this demonstrates that the flame undergoes violent oscillations to the left and right under the influence of sound waves. The reduced flame height indicates that the 3000 Hz frequency is less effective at promoting combustion than 2500 Hz. However, compared to the 1500 Hz effect, the flame tip is continuously stretched and bent, leading to a sharp decline in flame stability, as shown in Figure 10f.

3.3. Effects of Different Frequencies and Sound Pressure Levels on the Flame

As shown in Section 3.2, among the frequencies selected for the forward configuration, 1500 Hz exhibited the best flame suppression effect. When the frequency increased to 2500 Hz, the sound waves had a stimulating effect on the flame. It is evident that once the acoustic frequency reaches a certain threshold, the suppression effect peaks before transitioning to an excitation effect. Therefore, we further refined the optimal frequency range around 2000 Hz and adjusted the sound pressure level accordingly.

This optimization aims to determine the optimal attenuation conditions within a narrow frequency band. It is worth noting that the sound pressure level (SPL) naturally decreases with increasing distance from the sound source. Furthermore, unlike in ideal flame conditions, it is difficult to measure the SPL at the center of a pool fire in a pan during actual experiments. Therefore, when investigating the underlying mechanisms of acoustic fire suppression in the future, the SPL at the fire suppression location should be derived through attenuation calculations. When the sound pressure level was initially set to 20 dB, changes in the fluid flow were observed. Under the acoustic field parameters of 1800 Hz and 20 dB, as shown in Figure 11a, the deflection of the main flame body was slightly more pronounced compared to the 0 dB and 1500 Hz acoustic field. However, the suppression of temperature diffusion on both sides remained significant, yielding slightly better results. However, at 2000 Hz, as shown in Figure 11b, increasing the sound pressure level yields a more pronounced effect than no sound application. Although the flame body deflects in the opposite direction, fluid velocity data indicates the flame is being severely affected. The flame inclination angle increases, and oscillations begin to occur. When parameters increased to 2200 Hz 20 dB, as shown in Figure 11c, the acoustic wave’s effect on the flame was nearly equivalent to that at 1800 Hz 20 dB. This demonstrates that the suppression effect of high-frequency acoustic waves exhibits a curved growth within a certain range. The actual optimal operating parameters lie within this range, warranting further experimental validation for in-depth verification.

Plotting temperature and flow velocity results reveals that lateral sound sources exert significantly greater influence on temperature and fluid dynamics compared to frontal sources, as shown in Figure 12. At 1800 Hz 20 dB, the flame tilt angle indicates an effect nearly matching that at 2000 Hz 20 dB, as shown in Figure 12a. Under the lateral sound source, the flame body becomes more elongated, demonstrating that the combined effect of sound fields applied simultaneously on both sides is greater than that of a single frontal sound field. Further increasing the frequency to 2000 Hz 20 dB, as shown in Figure 12b, paradoxically reduces the disturbance to the flame morphology. The flame inclination angle becomes similar to that at 1800 Hz 20 dB and 1500 Hz 0 dB, with no significant change in temperature flow velocity. This further validates the conclusion that the suppression effect of high-frequency sound waves on flames exhibits a stepwise increase. As shown in Figure 12c, under 2200 Hz 20 dB conditions, the suppression effect was similar to that at 1800 Hz 20 dB, but the right side exhibited greater temperature diffusion, a steeper flame angle, more pronounced splashing phenomena, and enhanced oscillation effects.

3.4. Analysis of Flame Morphology Evolution Under Different Parameters

Observation of the normal combustion process reveals that the flame undergoes three significant morphological changes during a complete cycle: In the initial combustion phase, the flame gradually rises, stretches to a certain height, stabilizes, and begins sustained combustion. During the mid-combustion phase, the main body of the flame begins to surge toward the top in a wave-like pattern, with the tip transforming into a flame cluster. After the flame balls gradually separate from the main body, combustion enters the decay phase. The flame diminishes in size, fewer flame balls appear, and it reverts to a single main body state. The flame height decreases until it is completely extinguished. The influence of sound waves on the flame primarily arises from oscillations that disrupt the combustion morphology. Therefore, flame inclination angle is selected here to further analyze the aforementioned simulation data.

The flame inclination angle (θ) is defined as the angle formed between a line connecting the flame’s center point and its highest point within the region, and a horizontal line perpendicular to the flame’s highest point. The flame tilt angle is calculated manually based on simulated images of the visible flame. Kashiwagi and Newman [33] experimentally demonstrated that the flame tilt angle exerts a greater influence on the stability of the lower flame region than the external radiant flux. They further categorized unstable flames into three distinct forms. This indicates that changes in the flame tilt angle serve as a crucial metric for evaluating the effectiveness of parameters. As shown in Figure 13, the flame angle remains essentially constant below 1000 Hz. This phenomenon indicates that the aerodynamic disturbances generated by acoustic excitation below 1000 Hz are relatively weak and insufficient to cause significant, large-scale deflection or deformation of the flame. Only when the frequency exceeds a certain threshold does acoustic momentum exert a significant influence on the flame’s posture, leading to a noticeable change in the flame angle.

However, in actual kitchen grease fires, the combustion environment is often surrounded by multiple ignition sources such as oil storage containers and plastic packaging. Flames may even spread into exhaust ducts, further escalating the fire. This is a key reason for the easy spread of kitchen fires. Hirano et al. [34] studied flame propagation at different tilt angles using thin paper experiments, finding that flame spread speed continuously increased when the tilt angle ranged from 0 to 90°. Therefore, when selecting optimal working parameters for flame disturbance, one must also consider whether changes in flame tilt angle might increase the potential for fire spread, choosing the best option by balancing both factors.

As shown in Figure 14, under sound pressure level coupling, the flame inclination angle of forward sound waves exhibits no significant change but rather a decreasing trend. This phenomenon likely stems from the limited impact of forward sound sources on flame morphology. As seen in Figure 5, the flame inclination angle only undergoes significant change when the frequency reaches 2500 Hz. In contrast, when a lateral sound source is applied with sound pressure level, the flame angle is already affected by the sound waves as the frequency increases to 2200 Hz. This indicates that under certain sound source arrangement conditions, coupling frequency with sound pressure can enhance the disruptive effect on flame stability.

4. Experimental Study on the Effects of High-Frequency Sound Waves on Grease Fires

To validate the numerical simulation results, this chapter conducts experiments on the impact of high-frequency sound waves on the combustion stability of oil smoke fires. The primary objective of these experiments is to investigate the effects of varying parameters on flame stability. By applying acoustic fields within a defined parameter range to cooking oil smoke fires and comparing them with non-acoustic conditions, changes in flame images, center temperatures, and extinction times will be analyzed to determine the optimal acoustic field parameters for disturbance effects in practical applications.

4.1. Experimental Setup and Procedure

To ensure the full reproducibility of the experimental results, a schematic diagram of the laboratory setup and a photograph of the actual measurement configuration are provided in Figure 15. The experimental setup comprises three components: a cooking stove for oil smoke combustion, an acoustic extinguishing device, and a data acquisition system. A small household cast iron wok (11 cm diameter, 5 cm height) was selected as the cooking vessel, using common household rapeseed oil as fuel. Conducted in a dedicated laboratory at the university, the experiment was positioned away from doors and windows to minimize external airflow interference. Specific experimental conditions were recorded as follows: standard atmospheric pressure, room temperature of 13–15 °C, moderate indoor humidity of approximately 30–40%. The acoustic fire suppression device comprised a sound generator, power amplifier, loudspeaker, and gantry mounting frame. The sound generator produced specific acoustic signals transmitted to the power amplifier, which amplified the signals before delivering them to the loudspeaker for sound emission. Considering the relatively high maximum flame height of the oil pan, a forward configuration risks damaging the equipment. Therefore, a minimum clearance height of 60 cm was selected—the distance from the speaker plane to the oil pan’s ignition plane. In a lateral configuration, the distance from the midpoint of each side speaker to the midpoint of the oil pan’s ignition plane is 60 cm. An infrared thermal imaging camera captured the flame morphology and central temperature during combustion, while a stopwatch recorded the total duration of each burn. The equipment configuration and parameters for each component of the experimental setup are shown in

Table 3. Equipment configuration and parameters of experimental setup components.

Experimental System	Experimental Setup	Specifications	Function
Oil Fume Combustion System	Household Small Cast Iron Pot	Diameter: 11 cm Depth: 5 cm	Holds fuel rapeseed oil
	Heating Stove	Diameter: 16 cm Depth: 3 cm	For cast iron pots Continuous heat supply
Acoustic Interference System	Acoustic Interference Device	Sensitivity: 0~85 dB Frequency response: 80 Hz–20 kHz	Emit sound waves Flame Disturbance
Data Measurement System	Infrared Thermal Imager	±2 °C or ± 2% −20–150 °C, 100–550 °C	Capturing the combustion process
	Stopwatch		Calculating extinguishing time

. Based on the simulation results, the operating frequency range selected for this experiment is 1500–2200 Hz, for the following reasons: Although numerous laboratory studies have demonstrated that low-frequency sound waves can effectively suppress flames, from a practical engineering perspective, low-frequency sound waves have long wavelengths and strong penetrating power, which can easily induce structural resonance in building components and generate excessive low-frequency noise in enclosed indoor environments; therefore, they are not suitable for practical kitchen fire suppression applications. In addition, high-frequency acoustic fire suppression technology has been extensively studied in previous research, including the pioneering work by Viet and Tran at the University of Dammam on controlling oil fires using high-frequency acoustic excitation, as well as related patented technologies [35,36,37]. These studies have confirmed the feasibility and advantages of high-frequency sound waves in actual fire suppression scenarios, which aligns with the focus of this study on the effects of high-frequency sound waves on oil pan fires.

To simulate the conditions of a household frying pan, an open flame was used to continuously heat a cast-iron pan. The entire stove is made of heat-resistant clay; the bottom of the pan has a diameter of 13 cm, and the space beneath the grill is approximately 8.5 cm. The acoustic flame modulation device was designed by our team. The support frame is made of high-temperature-resistant steel and can be adjusted in height as needed. To ensure data accuracy, the speaker was secured to a cantilever until it was completely stable, ensuring that the angle of acoustic incidence remained consistent within the same control group. Specific parameter data are shown in

Table 4. Parameters of the acoustic flame disturbance device.

Laboratory Equipment	Manufacturer	Model	Parameters
Speaker	AIGO	X131	Sensitivity: 0~85 dB Frequency response: 80 Hz–20 kHz
Power amplifier	SAST	SU-8008	Power: 80 W Bluetooth connection
Infrared thermal imager	HIKMICRO	HIKMICROH16	Accuracy: ±2 °C or ±2% Operating temperature: −20–150 °C, 100–550 °C

.

The sound output of the loudspeaker is controlled by a power amplifier. The relative sound pressure level is expressed as a nominal sound pressure level (SPL) value, which is calibrated at a reference distance of 0.6 m from the loudspeaker in an anechoic environment (without a flame). Five nominal sound pressure levels were used: 20 dB, 30 dB, 40 dB, 50 dB, and 60 dB. Due to the intense nature of the pool fire, high temperatures, and acoustic flow disturbances in the vicinity, it was not possible to perform direct on-site sound pressure level measurements at the flame location. Therefore, the reported sound pressure level values represent the relative excitation intensity of the sound source, rather than the absolute sound pressure level at the flame. All comparisons between different frequencies and configurations were performed under the same sound source output settings to ensure the validity of the relative trend analysis. Direct measurement of the sound pressure level (SPL) at the flame location was not feasible in the present experimental setup due to the combined effects of thermal radiation, acoustic flow disturbances, and the risk of damaging the measurement microphone.

The SPL at the flame location, L_p, can calculate using the following engineering approach:

$$L_p=L_{p0}-A_{div}-A_{atm}$$

(20)

$$A_{div}=20 {\log }_{10}({\displaystyle \frac{r}{r_0}})$$

(21)

A_atm is the attenuation due to atmospheric absorption, calculated following the analytical method specified in ISO 9613-1, as a function of the acoustic frequency (1500–2200 Hz), air temperature (13–15 °C), relative humidity (30–40%) and atmospheric pressure (101.3 kPa) [38].

The use of theoretical SPL estimation based on source calibration and standard propagation models is a common practice in acoustic flame suppression research when in situ measurement at the flame location is impractical. This approach was also adopted in related studies [18,39].

This study used the source-level SPL settings as the control parameter, rather than the actual SPL at the flame location. While this approach does not allow for absolute quantification of the acoustic energy incident upon the flame, it is sufficient for comparative analysis under identical experimental conditions. Future work should employ high-temperature-resistant microphones to directly measure the flame-site SPL for more accurate modeling.

4.2. Effect of Different Sound Source Arrangements on Flames

The extinguished substance in this study is commercial rapeseed oil, which is used to simulate typical household cooking grease fires. Rapeseed oil is mainly composed of fatty acid triglycerides, with the primary components including oleic acid (C₁₈H₃₄O₂), linoleic acid (C₁₈H₃₂O₂), and palmitic acid (C₁₆H₃₂O₂). The general chemical formula for typical vegetable oil triglycerides can be expressed as C₅₅H₉₈O₆, which represents the dominant hydrocarbon structure involved in the combustion reaction. This experiment utilized the apparatus described above. During the experiment, 15 mL of rapeseed oil was placed in a cast-iron pot, ensuring the oil evenly covered the bottom of the pot. The procedure for the control group is as follows:

(1) Set up the experimental setup for the acoustic disturbance of oil in a pot over an open flame, as shown in Figure 15.

(2) Activate the acoustic disturbance device and verify that all system components are functioning properly. Use the Bluetooth-connected software to adjust the power settings, which is frequency sound wave generator. Ensuring that the power amplifier and speakers operate normally as parameters are modified. If no abnormal sounds are detected, place the rapeseed oil in the pan and ignite the alcohol block to begin heating.

(3) Record images of the flame combustion and temperature measurements under conditions without an acoustic field, serving as the initial control group.

(4) Adjust the experimental frequency to 1500–2200 Hz, dividing the range into four groups. Couple each group with one of five sound pressure level ranges (20–60 dB). Use an infrared thermal imager to record the experimental process until the oil is fully consumed and the flame extinguishes naturally. Select images of flame changes from three distinct stages and record the data.

(5) Compare the experimental images and data from each group to determine the optimal operating parameters.

(6) Repeat the above steps using both forward and lateral sound source configurations. To ensure experimental efficiency and accuracy, each set of experiments is repeated three times to minimize errors in the final data.

(7) Analyze the results to identify patterns in how flame morphology and extinguishing time vary with parameters, providing data support for subsequent research.

All experiments were conducted in a closed laboratory with controlled environmental conditions (temperature: 14 ± 1 °C, relative humidity: 35 ± 5%). Before the experiment began, the flame state of the oil pan fire without external acoustic disturbance was recorded, documenting the flame state and central temperature at each stage of normal combustion. To ensure repeatability, a standardized heating protocol was adopted. The cast-iron pot containing 15 mL of rapeseed oil was placed on a ceramic stove. The timer for combustion duration was started when the oil surface showed a stable, self-sustaining flame for more than 2 s, and was stopped when no visible flame was observed for 5 consecutive seconds. As shown in Figure 16a, when the oil temperature reached 415–420 °C, a flash fire occurred, and the flame became unstable. Timing commenced only after the flame had fully ignited and no sudden extinguishment occurred. Within the first minute, as shown in Figure 15b, the flame exhibited sustained growth within a certain height range. Fireballs continuously appeared at the flame tip, while the central flame temperature remained between 240 and 250 °C. After 1 min of combustion, the flame temperature increases, the flame height limit expands, the tip exhibits a wavy pattern without lateral oscillation, and the flame center temperature stabilizes between 340 and 360 °C, as shown in Figure 15c. The critical extinction temperature point ranges between 85 and 92 °C, as depicted in Figure 15d.

Initial control group data, as shown in

Table 5. Initial control group data.

	1	2	3	CV
Oil temperature at flash ignition	418.3 °C	419.9 °C	420.1 °C	0.24%
Time from flash to ignition	00.32.02	00.29.45	00.30.67	0.7%
(When sustained combustion reaches 1 min) Flame center temperature	244.7 °C	245.1 °C	247.9 °C	1.8%
(When continuously burning for 2 min) Flame center temperature	341.6 °C	346.3 °C	353.8 °C	4.4%
Temperature before flame extinguishes	84.1 °C	90.2 °C	91.4 °C	4.2%
Burning time	03.54.89	03.55.24	03.56.34	0.32%

, indicate that under no acoustic influence and with a fixed fuel quantity, the flame burned until complete extinction in approximately 3 min and 55 s. After 2 min of combustion, the flame gradually stabilized. At the 3 min mark, both flame length and temperature exhibited a stepwise decrease. Prior to extinction, the flame exhibited elliptical oscillations with a pronounced tendency to sway.

Arrange the forward sound source configuration by mounting the speaker on the beam directly above the stove, maintaining a minimum height of 60 cm between the speaker plane and the ignition plane. Following the same procedure as the silent test, start timing after the flash fire concludes and ignite the flame. Activate the sound field with frequencies set at 1500 Hz, 1800 Hz, 2000 Hz, and 2200 Hz, and sound pressure levels at 20 dB, 30 dB, 40 dB, 50 dB, and 60 dB. Record flame morphology changes, combustion duration, and center temperature for each setting.

Compared to the control group, the flame subjected to acoustic disturbance exhibited oscillatory fluctuations, burning in an unstable pattern. The combustion process shortened, and the temperature rise accelerated, with the flame temperature maintained between 255.4 and 297.8 °C at 1 min. During the mid-combustion phase, flame height decreased. The combustion state resembled that at 1 min without sound pressure, where a fireball formed at the flame tip and repeatedly detached from the main body during oscillations. At this stage, the central flame temperature ranged from 278.5 to 338.7 °C.

Based on the experimental results obtained when the sound field was set to 1500 Hz, as shown in Figure 17a–e indicate that increasing sound pressure levels enhances flame suppression, lowering the maximum central flame temperature and causing flame state disruption. Analysis of this phenomenon reveals that when the flame is disturbed by sound waves, the resulting acoustic flow accelerates air velocity near the flame. This increases oxygen concentration around the flame, promoting combustion. This manifests as a shortened combustion process and accelerated peak temperature attainment at the flame center. However, during the mid-to-late combustion stages, the flame oscillates with the surrounding air under the influence of sound waves, accelerating temperature flow. Simultaneously, the acoustic flow disrupts the flame morphology, generating unstable combustion patterns. Coupling sound pressure level with frequency can amplify this suppression effect. The average of three sets of combustion time data yielded a coefficient of variation ranging from 0.22% to 2.22%, confirming the reliability of the data. Comparing flame extinction times under acoustic field interference reveals that stronger sound pressure levels exert a greater influence on flame extinction duration. For detailed experimental data, see Appendix B.

After conducting the remaining three sets of experiments, comparing flame morphology, flame temperature, and total combustion duration when the sound pressure level was set within the range of 20–60 decibels. The following section contrasts flame changes under forward sound sources at different frequencies.

Compared to the 1500 Hz, 60 dB acoustic condition, the 1800 Hz, 60 dB sound field exhibited more pronounced flame suppression. The flame exhibited unstable combustion behavior, with temperatures at 1 min ranging between 240.1–251.9 °C and a decrease in peak central flame temperature. During the mid-combustion phase, flame height remained largely consistent. At 2 min, temperatures ranged from 305.2 to 328.7 °C, with the peak central flame temperature increasing. Significant temperature variations within different groups indicate further expansion of combustion instability, leading to a shortened combustion process. The average extinction time was 3.32 ± 0.1 min. At 2000 Hz, 60 dB, the flame temperature remained between 230.1 and 240.1 °C at 1 min, with the central flame temperature continuing its downward trend. As combustion stabilized, temperatures at 2 min ranged from 307.1 to 330.7 °C. Combustion instability continued to increase, reducing the average extinction time to 2.40.28. When frequency increased to 2200 Hz, the central flame temperature showed an overall downward trend: 242.1–244.5 °C at 1 min, and 251.9–253.1 °C at 2 min. The flame height remained largely stable throughout combustion with no significant variation, occasionally exhibiting undulating instability. The average extinction time extended to 4.05.36.

Analysis of the combustion duration data in Figure 18 indicates that flame extinction time decreases with increasing frequency, achieving optimal suppression at 2000 Hz. Beyond this frequency, combustion duration paradoxically increases. When frequency and sound pressure level interact, higher sound pressure levels shorten combustion duration. This effect is pronounced between 20 and 40 dB and moderates between 40 and 60 dB. As shown in Figure 19a–c, the flame center temperature decreases with increasing sound pressure level between 1500 Hz and 200 Hz. Analysis indicates that as sound waves suppress the flame more intensely, the surrounding airflow accelerates, leading to faster heat dissipation from the flame. At 2200 Hz, the temperature range narrows and stabilizes at a lower level, prolonging the flame extinction time. Therefore, it can be concluded that under the influence of a frontal sound source, the flame suppression effect increases with rising frequency and sound pressure, peaking at 2000 Hz and 60 dB. At 2200 Hz, the flame tends toward stability, with an extinction time even longer than under silent conditions.

Experimental results from the 1500 Hz sound field indicate that oblique sound sources exhibit effects similar to frontal ones. When the sound source is activated, the flame undergoes disturbance and begins oscillating, burning in an unstable state. Compared to the silent field and frontal sound source conditions, the combustion process continues to shorten, with flame temperatures maintained between 237.0 and 288.4 °C at 1 min. By the mid-burning stage at 2 min, the flame height decreased, and fireballs continuously detached from the main body at the flame tip. At this point, the flame center temperature ranged from 258.2 to 282.0 °C. The average combustion times are shown in Figure 20. Compared to the forward sound source, both the flame center temperature and total combustion duration decreased.

Analysis of flame variations under different parameter combinations for lateral sound sources yields conclusions consistent with forward sound sources. However, compared to forward sound sources, lateral sound source arrangements achieve optimal suppression effects at 1800 Hz. When frequencies increase further, such as at 2000 Hz and 2200 Hz, both flame center temperature and combustion duration indicate diminished suppression effects. As sound pressure level increases, the flame suppression effect grows in tandem with sound pressure, mirroring results from the frontal sound source. Under lateral sound source arrangements, sound pressure exhibits a steady, modest increase until reaching 50 dB, then stabilizes near its peak at 60 dB, as shown in Figure 21.

4.3. Influence of Different Acoustic Field Parameters on Flames

Combining the simulation results from Chapter 3 with the analysis of the above data, it is evident that sound waves possess the ability to disrupt flame stability through perturbation, with an optimal frequency range for this effect. The observed flame changes under applied sound fields align with simulation outcomes. Within specific parameter ranges, flame suppression first increases with parameter values, then levels off or even decreases after reaching a peak. Based on simulation analysis, the following conclusions can be drawn: Under a single sound field, excessive energy concentration can trigger localized over-disturbance, leading to a fueling effect. At high frequencies, the flame noticeably deflects toward one side, increasing diffusion on the deflected side. In contrast, a dual sound field distributes energy to both sides, avoiding excessive excitation in a single direction. This mutual compensation of deflection effects keeps the flame relatively centered, reducing the risk of fire spread to one side.

Comparing the two configurations, the forward-facing sound source achieves optimal suppression at 2000 Hz, 60 dB, while the side-facing sound source reaches peak effectiveness at 1800 Hz, 50 dB. This demonstrates that oblique sound waves from both sides are more effective at destabilizing the flame base. This effect correlates with the enhanced fluid velocity resulting from the superposition of sound waves in the dual-speaker arrangement. The increased airflow velocity around the flame caused by the sound waves weakens the flame boundary layer, making the flame more fragile. When dual sound fields act simultaneously from both sides, the boundary layer is weakened synchronously on both sides of the flame, creating a more uniform suppression effect with superior results.

The key parameters for acoustic fire suppression include frequency, sound pressure level, distance, and waveform modulation. These parameters should be adjusted based on the actual flame height, heat release rate, and fuel type. Frequency determines the disturbance mode and penetration, while sound pressure level determines the intensity of the acoustic force. The acoustic radiation force exerted by sound waves on the flame is given by the acoustic flux:

$$F={\displaystyle \frac{I}{c}}S$$

(22)

$$I={\displaystyle \frac{p^2}{\rho c}}$$

(23)

Methods such as focused beams, resonance amplification, and phased-array loudspeakers can be employed to expand the effective range of the technology. Currently, the maximum available frequency is 3000 Hz, which allows for effective control before the fire intensifies and spreads to the pipeline. Additionally, waveform modulation significantly affects fire extinguishing efficiency by enhancing periodic mixing and flow instability, warranting further in-depth research. The operational limits of this technology are primarily constrained by acoustic distance attenuation, thermal expansion of the flame, and environmental airflow disturbances. Intelligent systems have been shown to play a positive role in the initial detection of fires [40]. Koklu and Taspinar [41] applied five machine learning models to classify flame extinction status based on acoustic parameters, achieving the highest classification accuracy of 97.06% with the stacking method, demonstrating the feasibility of AI-assisted decision support in acoustic fire suppression systems. Ivanov and Stankov [42] presented an intelligent acoustic extinguishing platform that integrates Deep Neural Network (DNN)-based flame and smoke detection with an acoustic fire extinguisher, enabling immediate fire suppression upon positive detection without unnecessary time delay. Wilk-Jakubowski et al. [43] developed a high-power acoustic fire extinguisher integrated with a deep neural network (DNN)-based flame detection module, demonstrating automatic activation upon flame detection without the need for traditional sensors. Future research will focus on the synergistic effects of artificial intelligence and acoustic fire suppression. The integration of intelligent systems will make the application of these parameters more precise and efficient.

5. Conclusions

• Acoustic waves can effectively destabilize oil pan flames. The suppression effect increases first and then decreases with the increase in frequency, showing a nonlinear law.
• On the basis of the optimal frequency, increasing the sound pressure level can significantly enhance the suppression effect, but it has saturation characteristics.
• The bilateral oblique sound source acts on both sides of the flame root at the same time, which can realize uniform suppression and compensate for the high-frequency deflection effect of a single sound field, with higher efficiency.

In summary, acoustic fire suppression technology demonstrates significant intervention efficacy against commercial kitchen fires, with its operational mechanisms jointly regulated by frequency, sound pressure level, and source configuration. The optimal parameter combination identified in this study is 1800 Hz at 50 dB, with a preferred bilateral oblique arrangement. Although this study is aimed at kitchen oil fires, the conclusions and parameter laws can also be applied to similar liquid fuel fires in enclosed chambers. These findings provide theoretical foundations and data support for the engineering application of acoustic fire suppression devices.