Two-dimensional Numerical Model Research Articles

Numerical Investigation on the Effects of Stingray Flapping Amplitudes and Counts on Self-Burial Performance

Marine benthic organisms, including stingrays, utilize sediment for self-burial, providing an effective strategy for concealment and protection against currents. This study investigates the effects of stingray flapping amplitude and count on self-burial performance using a two-dimensional numerical model, and two self-burial strategies are proposed. The first involves a single large-amplitude flap, achieving high sand transport efficiency but demanding higher transient power. The second uses multiple medium-amplitude flaps, results in a greater burial depth with lower peak power requirements. The analysis reveals that effective self-burial is dependent on raising an adequate quantity of sand and ensuring its deposition on the stingray body. Vortex dynamics play a crucial role, with appropriate vortex strength and sand volume needed for complete body coverage. Furthermore, this study underscores the potential of these findings to inform the design of bio-inspired marine robots capable of self-burial, offering novel insights into the mechanics underlying stingray self-burial and its broader implications for underwater technology development.

Simple model, complex strata: 2-D numerical forward model analysis of heterogeneity and controls in submarine fan systems

ABSTRACT Turbidite strata stacking patterns are interpreted as products of either allogenic forcing, autogenic processes, or a combination of both. However, the relative influence of each remains difficult to constrain. Here, a simple reduced-complexity 2-D numerical forward model (Lobyte2D) simulates gravity flows passing down a slope, running out, and depositing across a flat basin floor. Flows can erode, bypass, or deposit, cumulatively producing a variety of entirely autogenic retrograding, aggrading, and prograding strata with stratigraphic completeness values ranging from less than 1% to around 40%. Complexity of modeled strata is measured with a spatial entropy metric, and sensitivity analysis indicates that grain-size and flow-acceleration parameters are the key controls on stratal complexity: larger grain sizes are associated with deeper erosion, which makes more rugged topography, and higher values of flow acceleration make flow speeds more sensitive to topography, which triggers localized deposition and/or erosion. Such complexity produced by a simple two-dimensional numerical forward model suggests that even more complex behavior is likely in natural systems, and this should be reflected in outcrop and subsurface interpretations. However, comparison of geometries in chronostratigraphic and cross-section plots of modeled strata shows that, due to a variety of cryptic bypass and erosion surfaces, stacking trends visible in chronostratigraphic plots are much more difficult to detect in outcrop and subsurface cross sections and vertical sections. These insights suggest that many outcrop and subsurface interpretations of submarine-fan strata, particularly sequence stratigraphic interpretations, may be missing substantial complexity, and underestimating the uncertainty inherent in limited data.

Deep dehumidification characteristics of a silica gel coated cross-flow heat exchanger with a circulating blowing loop

Solid desiccant dehumidification systems offer an effective and energy-efficient alternative for deeply dehumidifying air. Desiccant coated heat exchangers (DCHEs), a type of solid desiccant dehumidification system, have garnered attention due to their ability to eliminate adsorption heat and thereby enhance dehumidification capacity, unlike desiccant wheels that perform adiabatic dehumidification. This study introduces a silica gel coated cross-flow DCHE equipped with a circulating blowing loop for deep dehumidification applications. A two-dimensional numerical model is developed to simulate deep dehumidification behavior of the DCHE. Parametric analysis is performed to explore the dehumidification characteristics under various conditions, setting different volume fractions ω of blowing loop air. Results indicate that the DCHE can effectively reduce the moisture content of humid air to a deep dehumidification threshold of 6.2 g/kg. Additionally, dehumidification capacity can be increased by incorporating a circulating blowing loop. The influence of process air velocity, cooling air velocity, process air humidity ratio, regeneration air temperature, and air channel length on the dehumidification performance are analyzed, focusing on metrics such as minimum outlet air humidity ratio Ya1,ad,out_min, effective deep dehumidification time teff, moisture removal capacity MRC and dehumidification coefficient of performance DCOP. It is determined that the regeneration air temperature should be raised to 80 °C to ensure deep dehumidification, and an air channel length of 0.2 m is optimal.

Research on the Mechanism of Femtosecond Laser Ablation Concave-Convex Microstructure Transformation.

The phenomenon of femtosecond laser ablation of convex structures from the bottom to the top is interesting. In this study, AZ31B magnesium alloy was used as the substrate to analyze the impact of the laser pulse energy and scanning speed on the morphology of concave-convex microstructures. Subsequently, a unified two-dimensional numerical model incorporating solid, liquid, and gas phases was established, and combined with experimental data, the mechanism and formation process of concave-convex transformation in magnesium alloy under laser ablation were revealed. The results indicate that the transition from concave to convex structures is significantly influenced by the laser scanning speed, whereas the laser pulse energy primarily affects the shape and size of the convex structures. During the ablation process, molten material is expelled and gradually accumulates on both sides of the ablation groove under the action of the recoil pressure. During cooling, the molten material at both ends of the groove merges to form protrusions under the combined effects of internal negative pressure, gravity, and Marangoni forces. Moreover, this method of femtosecond laser ablation for generating convex structures deviates from the traditional single-texture approach to concave structures, potentially broadening the application of laser composite processing surfaces.

The decoupled effect of oxidizer mass flux and combustion pressure on turbulent combustion characteristics of finite-length solid fuel

Combustion pressure and oxidizer mass flux are key factors influencing the performance of hybrid rocket motors. In the present work, the decoupled effects of oxidizer mass flux and combustion pressure on turbulent combustion characteristics were investigated. First, a finite-length combustion theory of fuel grains was developed to identify the main sensitive parameters, including flame structure and the recirculation zone, which affect the combustion of solid fuel. Then, a two-dimensional transient numerical model, based on the coupling characteristics of the heat transfer process in solid fuel grains and dynamic combustion flow, was developed and validated through fire experiments under a wide range of inflow conditions. A quantitative analysis was conducted to assess the impact of flame structure and the recirculation zone on the dynamic combustion characteristics of finite-length solid fuel. The results showed that the fuel characteristic temperatures of the front and central parts are negatively correlated with flame height, while the characteristic temperature of the rear part is positively correlated with the recirculation zone temperature. Both the mass flow rate and combustion pressure are negatively correlated with flame height and positively correlated with the temperature of the recirculation zone. Compared to combustion pressure, the influence of mass flow rate is more significant. The transient processes of the solid fuel regression rate under different inflow conditions exhibit a consistent tendency: an increase in mass flow flux and combustion pressure results in faster attainment of peak regression rate and stabilization.

Mechanical Characteristics of Deep Excavation Support Structure with Asymmetric Load on Ground Surface

The purpose of this paper is to capture the mechanical response of the support structure of deep excavation subject asymmetric load. A two-dimensional (2D) numerical analysis model was established by taking a pipe gallery deep excavation subject to asymmetric load as an example. The numerical analysis results were in good agreement with the measured data, thus verified the validity of the numerical model. On this basis, the stress and displacement of support structure caused by the change in foundation asymmetric load were studied. According to the numerical results, horizontal displacement of the diaphragm wall (DW) was dominant, and the maximum horizontal displacement of the DW was 7.54 mm when the deep excavation was completed. With the increase in asymmetric load, the left wall displacement continued to increase, while the displacement of the right DW continued to decrease, and the maximum horizontal wall displacement occurred near the excavation face. The DW was the main bending component, and the maximum wall bending moment when the deep excavation was completed was 173.5 kN·m. The maximum wall bending moment increased with the increase in asymmetric load, and the maximum wall bending moment on the left of the deep excavation was greater than that on the right. The inner support sustained the main component of axial force, with the axial force peaking at 1051.8 kN when the deep excavation was completed. The axial force of the inner support increased with increasing the asymmetric load, and the axial force of the second inner support was obviously greater than that of the first inner support. This research has a positive effect on the design and optimization of deep excavation support structure subject to asymmetric load on ground surface.

Study of self-assembly between two magnetic particle chains in magnetorheological fluids

The self-assembly behavior of individual magnetic particles in magnetorheological liquids has been extensively analyzed in previous studies, but the self-assembly behavior between chains of magnetic particles has rarely been investigated. In this paper, the self-assembly mechanism between chains of magnetic particles is investigated using simulation and experimental methods. Firstly, a two-dimensional coupling simulation numerical model of particle-magnetic field-flow field is constructed by analyzing the magnetic force, fluid force and contact interaction force between magnetic particles. The accuracy of the simulation model was verified through experiments. Then the attraction and repulsion dividing line of two magnetic particle chains are analytically calculated by simulation combined with parametric scanning. And analyze the law of the influence of the number of magnetic particles between magnetic particle chains on the attraction and repulsion dividing line. Finally, the different self-assembly behaviors that occur when magnetic particle chains are in different regions of attraction and repulsion between them are experimentally verified. Thus, this study provides new insights into the self-assembly mechanism between chains of magnetic particles in magnetorheological fluids in terms of the effect of the number of magnetic particles on the attraction–repulsion dividing line.

Pressure oscillation and suppression method of large-aspect-ratio solid rocket motors

A two-dimensional large eddy simulation numerical model is proposed to study the transient vortex flow and pressure oscillation of a large-aspect-ratio solid rocket motor. The numerical model is validated through experimental data, finite element analysis and cumulative error analysis. The numerical simulations are executed to obtain the characteristics of the vortex-acoustic and pressure oscillation. The results show that the burning surface regression decreases the motor aspect ratio, increasing the corresponding natural frequency from 260 Hz to 293 Hz. The pressure oscillation phenomenon is formed due to the vortex-acoustic coupling. Decreasing the corner vortex shedding intensity shows negative effects on the dimensionless amplitude of the pressure oscillation. The head cavity without the injection can decrease the vortex-acoustic coupling level at the acoustic pressure antinode. The modified motor with head cavity can obtain a lower dimensionless oscillating pressure amplitude 0.00149 in comparison with 0.00895 of the original motor. The aspect ratio and volume of the head cavity without the injection have great effects on the pressure oscillation suppression, particularly at the low aspect ratio or large volume. The reason is that the mass in the region around the acoustic pressure antinode is extracted centrally, reducing the energy contribution to the acoustic system. With the volume increasing, the acoustic energy capacity increases.

Kinematics and hydrodynamic performance of zebrafish C-type maneuvers: A comparison of two- and three-dimensional simulations

Two-dimensional (2D) and three-dimensional (3D) numerical models are commonly employed to investigate the kinematic and hydrodynamic characteristics of fish maneuvers. In this study, we captured the posture characteristics of zebrafish during C-type maneuvers using high-speed photography and constructed a midline curvature model via the tandem principal characteristics method, which exhibited a “double peak and single valley” structure. Based on this curvature model, self-propelled simulations were conducted using the immersed boundary method with adaptive mesh refinement. The results showed that, under identical deformation conditions, the 2D simulation exhibited a 16.8% higher centroid velocity, 6.1% greater overall angular velocity, and an 11.9% larger turning angle compared to the 3D simulation. This discrepancy is primarily due to the 2D model’s inability to accurately represent the fish body’s mass distribution and force characteristics, resulting in artificially elevated performance. Nevertheless, 2D simulations remain applicable for studying the propulsion performance of fish with elongated cross-sections and large fin areas. Comparison between the simulated and real motion performance reveals that, under the self-propelled computational model, both 2D and 3D numerical simulations consistently capture the qualitative motion patterns. The quantitative results also reflect the actual swimming performance of the fish within an acceptable margin of error.

Heat transfer characteristics of methane-air premixed jet flames with flat/hemispherical walls

Abstract The in-depth study of the mutual coupling between the flame and the wall can significantly enhance the efficiency of actual combustion devices. A two-dimensional numerical model of the heat transfer characteristics of methane-air premixed jet flames with flat/hemispherical walls has been developed. An examination of the effects of wall shape on the heat transfer characteristics of methane/air flames was conducted as a function of the equivalence ratio (ϕ=0.9-1.5), the mixture Reynolds number (Re=300-800), and the burner-to-plate distance (H/d=1-6). As the equivalence ratio and Reynolds number increase, the flame temperature increases on the surface near the wall, and the temperature near the flame centerline is higher under the influence of a hemispherical wall than it is under the influence of a plate. In addition, the wall's heat flux increases as both the equivalence ratio and the Reynolds number increase. It is observed that the heat flux of the hemispherical wall is greater than that of the flat plate near the stagnation point, whereas it is smaller at a distance from the stagnation point. Due to the burner-to-plate distance, thermal efficiency is maximized when the flame premixed cone contacts the impact surface, which is the desired condition for optimal performance. Due to different operating conditions, the efficiency of heat transfer is always higher under the action of a flat plate than under the action of a hemispherical wall.

Mixing and ignition characteristics of laser-induced solid fuel pyrolysis under wide-range variation of crossflow

A theoretical model has been developed to elucidate the thermal initiation laser ignition process of solid fuel. This model comprehensively explores both the heating process of solid fuel and the mixing process of gas-phase pyrolysis gases with crossflow conditions. A two-dimensional transient numerical model has been established and validated through experimental investigations across a spectrum of crossflow fluxes (ranging from 20 to 150 kg/m2/s) and pressures (ranging from 0.15 to 1.83 MPa). The study involves a quantitative analysis of the influence of key macroscopic parameters, including laser energy levels, combustion pressure, and flow rates, on essential mixing characteristics. The solid fuel heating process is segmented into two distinct phases: a preheating phase and a pyrolysis phase, with the latter exhibiting considerably higher heating rates. Notably, a direct linear relationship is observed between the laser energy levels and the rate of heating. It reveals that the dynamics of pyrolysis product mixing are primarily determined by the solid fuel pyrolysis rate and the prevailing crossflow conditions. The dimensionless penetration depth, which measures the extent of pyrolysis product mixing with crossflow, demonstrates an exponential positive correlation with pyrolysis rates and oxidizer flow rates while displaying a negative correlation with combustion pressure. This analysis includes the quantitative determination of proportionality constants and power exponents within the correlations. Finally, the study delves into the intricate interplay between crossflow conditions and the criteria for laser ignition, specifically ignition temperature and the flammable lower limit.

Exploring ice melting dynamics in beverageware

The ice-melting process inside water-filled beverageware is studied experimentally, theoretically, and numerically. An ice body is immersed in warm fresh water in different-shaped glass cups. The ice body eventually melts due to heat transfer from the warmer fluid to the ice. As the ice melts and cools the surrounding fluid, it promotes complex time-dependent laminar flows characterized by a vertical downward jet beneath the ice body and rotating convective cells in the fluid domain. Experimental results obtained through dye visualization, particle image velocity, and local temperature measurements show that the size and number of the rotating cells, jet velocity, and melting time of the ice depend significantly on the shape of the beverageware. A one-dimensional analytical solution reproduces qualitatively the hydrodynamical and thermal boundary layers close to the ice body. Further, a two-dimensional numerical model correctly captures the main features of the experimental observations.

Double subduction controls on long-lived continental tectonics and subcontinental mantle temperatures

The complexities of convergent margins commonly include the interactions of subduction zones, with many geological records of “double” subduction. Here, we build two-dimensional numerical models to explore the evolution of complex subduction systems by systematically testing single and inward-dipping double subduction beneath a continental upper plate and the impact of continental collision on these systems. When compared to single subduction models, the inward-dipping double subduction shows hindered trench migrations and larger volumes of upwelling mantle enhanced by excess sinking slab mass. Double subduction draws larger volumes of hotter mantle beneath the continent in an area much broader than the marginal basins of single subductions, contributing to subcontinental heating by ∼200 °C. As collision jams one margin of a double subduction system, the other margin follows the evolution of migrating single subduction zones, although characterized by persisting higher mantle temperatures and strong upwellings, inherited from the double subduction stage, and large-scale upper plate extension. The modeling outcomes are compared to scaling arguments to test the viability of the mechanism proposed for tectonics of the Cenozoic South China Sea and Neoproterozoic Yangtze Block (southeastern China), where the inward-dipping double subduction provides a context for protracted large-scale continental extension, hotter subcontinental temperatures, and channeled mantle flow not easily reconciled with the dynamics of single subduction zones.

Simulation of hydrodynamic changes and salinity intrusion in the lower Vietnamese Mekong Delta under climate change-induced sea level rise and upstream river discharge

Saltwater intrusion results from a complex interaction of many factors, such as global climate change, extreme climate patterns and overuse of natural resources. Currently, the Vietnamese Mekong Delta (VMD) faces great challenges from saltwater intrusion. This situation is very serious, tends to increase in scope and scale, directly threatens life and property, and negatively impacts the sustainable economic development of this area. The objective of this study is to evaluate and predict changes in the hydrodynamic regime and salinity intrusion in the Lower VMD (from Can Tho and My Thuan to estuaries and coastal regions) caused by anticipated increases in sea level and the decrease in river discharge from upstream, using a two-dimensional numerical model. The research includes many scenarios, such as a baseline scenario in February 2020 and eight sea level rise (SLR) scenarios ranging from 14 cm to 58 cm in 2050 and 2100. The simulation findings indicate that when the sea level rises, there is a significant increase in salinity and saltwater intrusion length, resulting in increased wave height, decreased current speed, and a decline in upstream discharge of the rivers. The work has enhanced our comprehension of the impact of these hydrodynamic alterations on saltwater incursion and riverbank erosion.

Active and passive salt diapirs: a numerical study

Summary Salt diapirs dominate the structure in many sedimentary basins and control the preservation and migration of hydrocarbon. The formation of salt diapirs generally falls into two endmember models: active (up-building) and passive (down-building) diapirism. In the active model, salt diapirs rise from salt buoyancy to pierce through the sedimentary overburden, whereas in the passive model, salt diapirs result from differential loading of sediments during deposition. These endmember models are mostly conceptual or kinematic, the mechanics of active and passive diapirism, and their relative roles and interactions in the formation of salt diapirs, remain uncertain. Here, we use two-dimensional high-resolution numerical models to investigate the primary factors and critical conditions for active and passive diapirism. Our results indicate that it is improper to use driving mechanisms to classify salt diapirs, because the buoyancy-driven active salt diapirism involves differential loading, while the passive diapirism requires salt buoyancy. The rise of salt diapirs is more sensitive to the effective viscosity of the overburden than to the salt viscosity. Stiff overburdens could prevent the rise of salt diapirs, but they could be pierced by salt diapirs if plastic yield of the overburden is allowed. During deposition, the coupled salt-sediment deformation, driven by both salt buoyancy and differential loading of sediments, can lead to various diapiric salt structures and minibasins. Regional tectonic stress generally promotes salt diapirism by enhancing strain weakening of salts and overburdens. We suggest that the classification of active and passive salt diapirism is an oversimplification in most cases. We propose a general model of the formation of salt diapirs that usually begins with dome initiation driven by salt buoyancy, followed by syndepositional down-building controlled by sedimentation and differential loading, and ends with canopy formation when sedimentation stops.

Numerical analysis of tapered wick configurations in cylindrical heat pipes

Heat pipes are essential devices in thermal engineering due to their ability to transport heat with remarkable efficiency. They are crucial for heat distribution and control in a variety of applications, such as aircraft systems and electronics cooling. This paper presents a numerical analysis on the performance characteristics of a cylindrical heat pipe featuring a novel tapered wick geometry. A two-dimensional numerical model incorporating non-Darcian liquid transport is formulated and solved at steady state. To investigate the effect of the tapered wick geometry, major characteristics such as axial distribution of vapor pressure, liquid pressure and wall temperature are evaluated and analyzed. Based on the direction of taper, three configurations viz., THPE (Tapering wick heat pipe from evaporator end), THPC (Tapering wick heat pipe from condenser end) and THPA (Tapering wick heat pipe at adiabatic center) are studied. A novel non-dimensional parameter termed as Temperature Pressure Index (TPI) is proposed to signify the extent to which an HP can reduce its evaporator temperature relative to increase in pressure drop. Initially a parametric study is performed. Results showed a superior performance of THPE and THPC designs over THPA in terms of output parameters. The numerical findings demonstrate that when compared to a uniform wick heat pipe, both the THPE and THPC configurations exhibit a notable enhancement of 58 % in effective thermal conductivity under a maximum load of 455 W. Further, by conducting a variance-based (Sobol) sensitivity analysis, the THPC design is found to be marginally better as compared to THPE. Finally, the BOBYQA optimization algorithm is employed to find optimum taper angle for THPC design that maximizes the output parameters. The optimum taper angle of the wick is found to be 0.0372° for minimum average evaporator temperature in the present case. For this optimized THPC design, a reduction in average evaporator temperature of 6.51 K is obtained compared to uniform wick heat pipe. Additionally, the obtained TPI value is about 1.86 times than that of a uniform wick heat pipe.

Behaviour of liquefaction for Darbandikhan Dam consequence impact of seismic load

Abstract Embankment dams are at risk of damage due to the impact of earthquakes. Rock-filled dam structures are especially vulnerable to earthquakes. To assess the impact of earthquakes, it is important to study liquefaction, which is a significant factor affecting the dynamic behavior of the dams. This study aims to determine the probability that the Darbandikhan rock-fill dam on the (Sirwan- Diyala) river in Iraq may liquefy using two-dimensional numerical modelling and finite element analysis, which is located in Alsulamaniyah north of the capital of Iraq, Baghdad 230 km. The simulation results are compared in terms of different water levels 434, 472, and 485m.a.s.l., different peak accelerations of the earthquakes (0.02, 0.04, 0.06 and 0.08) and different duration of earthquakes 25, 50, 75, and 100sec. As, a result the liquefaction zone increases with increasing upstream water level, acceleration, and duration. So, when the maximum operation water level was 485 m, the percentage of liquefaction area was 20% at peak acceleration of 0.08 g, and a duration of 100s. After that, the liquefaction zone decreases with a decreasing upstream water level of 472 m, for the same acceleration and duration of up to 12.5%. However, this percentage when the minimum water level of 434 m, acceleration 0.08g, and duration 100s, decreases by 65% from the liquefaction zone at elevation 485 m.a.s.l.

High-dimensional multi-objective optimization of coupled cross-laminated timber walls building using deep learning

This paper introduces a deep learning-based multi-objective optimization framework, applying for advancing the design of the Cross-Laminated Timber Coupled Wall system. While traditional optimization methods often struggle with the curse of dimensionality in high-dimensional problems, the approach proposed in this study employs an autoencoder to effectively reduce the dimensionality of the design space. Subsequently, a neural network establishes a mapping between input variables and latent spaces, with another neural network forming the crucial link between these latent variables and the output responses. The proposed framework is integrated into the design process of a 20-story Cross-Laminated Timber Coupled Wall system, where uncertainties inherent in connection elements are systematically addressed to optimize the structural parameters. The non-dominated sorting genetic algorithm-II is utilized to estimate optimal design variables by minimizing three conflicting objective functions, thereby generating a Pareto front. This optimized design is then bench-marked against three deterministic models with varying coupling beam shear strengths. A two-dimensional numerical model developed in OpenSees facilitates nonlinear time history analysis using 50 bi-directional ground motion records, representative of the seismicity of Vancouver, Canada. The results of this study not only highlight the efficacy of the deep learning-based framework in enhancing the structural integrity and resilience of high-rise timber structures in seismic regions but also significantly contribute to the evolution of computational approaches in structural engineering.

Numerical investigations of GRS wall performance with tiered configurations

Geosynthetic Reinforced Soil (GRS) walls have become increasingly popular as a result of their numerous advantages. In some cases these structures are constructed in a multi-tiered configuration, which makes their behavior more complicated. Nevertheless, the response of the multi-tiered walls is insufficiently explained by the existing design manuals and literature studies. This paper investigated the performance of GRS walls with multi-tiered configurations using two-dimensional (2D) finite element numerical models. This study compares the performance of two-tiered GRS walls with simple GRS walls (single-tiered). It also examines the impact of offset distance, backfill strength properties, and reinforcement parameters (vertical spacing and reinforcement length) on horizontal deformations and reinforcement tensile loads in two-tiered GRS walls. Adopting a multi-tiered configuration for GRS walls can significantly reduce both the horizontal wall deformation and the maximum reinforcement tensile loads. Additionally, the critical offset distance found in this study is significantly smaller than that recommended by the Federal Highway Administration (FHWA) guidelines. The finite element results also demonstrate that using high-quality backfill soil can minimize interaction between lower and upper tiers. Using a uniform reinforcement length of 0.6H for both tiers significantly reduces horizontal deformation compared to the FHWA recommendation of 0.6H and 0.35H for the lower and the upper tier respectively.

Simulation and Experimental Investigation of the Effect of Pore Shape on Heat Transfer Behavior of Phase Change Materials in Porous Metal Structures.

With the gradual increase in energy demand in global industrialization, the energy crisis has become an urgent problem. Due to high heat storage density, small volume change, and nearly constant transition temperature, phase change materials (PCMs) provide a promising method to store thermal energy. In this work, we designed and fabricated three kinds of porous metal structures with hexagonal, rectangular, and circular pores and explored the phase change process of PCMs within them. A two-dimensional numerical model was established to investigate the heat transfer process of PCMs within different shapes of porous metal structures and analyze the influence of heat source location on the thermal performance of the thermal storage units. Visualization experiments were also carried out to reveal the melting process of PCMs within different porous metal structures by a digital camera. The results show that paraffin in a porous metal structure with hexagonal pores has the fastest melting rate, while that in a porous metal structure with circular pores has the slowest melting rate. Under the bottom heating mode, the melting time of the paraffin in porous metal structures with hexagonal pores is shortened by 18.6% compared to that in porous metal structures with circular pores. Under the left heating mode, the corresponding melting time is shortened by 16.7%. These findings in this work will offer an effective method to design and optimize the structure of porous metal and improve the thermal properties of PCMs.