With the rapid adoption of renewable energy sources, the domestic solar photovoltaic power generation industry has experienced significant growth. Solar inverters, which are critical components in converting direct current (DC) from solar panels to alternating current (AC) for household use, must efficiently manage thermal dissipation while minimizing the aerodynamic noise generated by their cooling systems. This noise, primarily originating from axial fans used for heat dissipation, can impact user comfort and environmental noise pollution. In this article, we investigate the aerodynamic noise characteristics of a solar inverter cooling system using a hybrid approach that combines computational fluid dynamics (CFD) and computational aeroacoustics (CAA). We validate our numerical model through experimental measurements and propose design modifications to reduce noise levels, emphasizing the importance of pressure fluctuations on fan blades and the influence of intake structures. Throughout this discussion, we will repeatedly reference solar inverters to highlight their central role in sustainable energy systems.
The cooling systems in solar inverters typically employ axial fans to maintain optimal operating temperatures by forcing air through heat sinks and electronic components. However, these fans can generate substantial aerodynamic noise due to unsteady flow phenomena, such as vortex shedding and pressure pulsations. Previous studies have often focused on isolated fans without considering the surrounding enclosure effects, leading to inaccuracies in noise prediction. Our work addresses this gap by incorporating the full inverter geometry into the simulation, ensuring a more realistic assessment of noise propagation. The integration of solar inverters into residential areas necessitates low-noise operation, making this research highly relevant for product design and optimization.
To analyze the flow field, we use the RNG k-ε turbulence model, which is suitable for complex flows with high strain rates and curved walls commonly found in solar inverter cooling systems. The governing equations for turbulent kinetic energy \(k\) and dissipation rate \(\epsilon\) are as follows:
$$\frac{\partial (\rho k)}{\partial t} + \frac{\partial (\rho k u_i)}{\partial x_i} = \frac{\partial}{\partial x_j}\left[\left(\mu + \frac{\mu_t}{\sigma_k}\right) \frac{\partial k}{\partial x_j}\right] + G_k – \rho \epsilon$$
$$\frac{\partial (\rho \epsilon)}{\partial t} + \frac{\partial (\rho \epsilon u_i)}{\partial x_i} = \frac{\partial}{\partial x_j}\left[\left(\mu + \frac{\mu_t}{\sigma_\epsilon}\right) \frac{\partial \epsilon}{\partial x_j}\right] + C_{1\epsilon} \frac{\epsilon}{k} G_k – C_{2\epsilon} \rho \frac{\epsilon^2}{k}$$
where \(\rho\) is the fluid density, \(t\) is time, \(u_i\) are the velocity components, \(\mu\) is the dynamic viscosity, \(\mu_t\) is the turbulent viscosity, \(G_k\) represents the generation of turbulent kinetic energy due to mean velocity gradients, and \(\sigma_k\), \(\sigma_\epsilon\), \(C_{1\epsilon}\), \(C_{2\epsilon}\) are model constants. The RNG model includes additional terms to improve accuracy for swirling flows, which are prevalent in axial fans within solar inverters.
For acoustic prediction, we employ a hybrid method that extracts pressure fluctuations from the CFD simulation and uses them as dipole sources in an acoustic finite element model. The Ffowcs Williams-Hawkings (FW-H) equation is adapted for enclosed spaces, accounting for reflections and scattering by the inverter housing. The sound pressure level (SPL) in decibels (dB) is calculated as:
$$L_p = 10 \log_{10}\left(\frac{p_{\text{rms}}^2}{p_{\text{ref}}^2}\right)$$
where \(p_{\text{rms}}\) is the root-mean-square sound pressure and \(p_{\text{ref}} = 20 \mu\text{Pa}\). The overall A-weighted SPL is obtained by summing contributions across frequency bands, relevant for evaluating human perception of noise from solar inverters.
We begin by simplifying the geometric model of a typical solar inverter to facilitate meshing and simulation. The original design includes intricate details such as bolt holes and cable ports, which are removed without affecting the overall flow dynamics. The fluid domain is divided into stationary and rotating regions, with the latter enclosing the axial fans. Key parameters of the fans are summarized in Table 1.
| Parameter | Value |
|---|---|
| Number of Blades | 7 |
| Fan Outer Diameter | 56 mm |
| Tip Clearance | 3 mm |
| Rotational Speed | 9,000 rpm |
| Base Frequency (f) | 1,050 Hz |
The mesh is generated using unstructured tetrahedral elements, with refinement near the fans and heat sinks to capture pressure gradients. The total mesh count is approximately 10.7 million elements, ensuring grid independence. Boundary conditions include pressure inlets and outlets set to atmospheric pressure, no-slip walls for solid surfaces, and rotating walls for fan blades. The multiple reference frame (MRF) approach is used to simulate fan rotation at 9,000 rpm. Steady-state simulations are performed first, followed by unsteady Reynolds-averaged Navier-Stokes (URANS) calculations with a time step of \(1 \times 10^{-5}\) seconds for 800 iterations.
Flow field analysis reveals significant vortices near the blade tips, particularly at the leading edge. These vortices cause pressure fluctuations on the blade surfaces, which are primary sources of aerodynamic noise in solar inverters. The static pressure distribution on the suction side of the fan blades shows regions of high negative pressure, correlating with elevated sound pressure levels in acoustic simulations. We observe that the pressure pulsation magnitude directly influences noise radiation, as described by the dipole source strength in aeroacoustics. The relationship between pressure fluctuation \(\Delta p\) and sound power \(W\) can be approximated by:
$$W \propto \rho_0 c_0^{-3} (\Delta p)^2 A^2 f^2$$
where \(\rho_0\) is air density, \(c_0\) is the speed of sound, \(A\) is the surface area, and \(f\) is frequency. This underscores the importance of minimizing pressure variations in solar inverter fans.
For acoustic modeling, the fan blades are segmented into equivalent dipole sources, and the acoustic domain is discretized using finite elements with a perfect matched layer (PML) to truncate the simulation space. The mesh size is 10 mm, yielding 1.2 million elements and a maximum resolvable frequency of 5,666 Hz, which covers the fundamental and harmonic frequencies of the fan. Field points are arranged on a hemispherical surface with a 2-meter radius, following standardized measurement procedures for solar inverters. The sound power level \(L_W\) is computed from the SPL measurements using:
$$L_W = L_p + 10 \log_{10}\left(\frac{S_2}{S_0}\right) + C_1 + C_2 + C_3$$
where \(S_2\) is the surface area of the hemisphere, \(S_0 = 1 \text{ m}^2\) is the reference area, and \(C_1, C_2, C_3\) are correction factors for environmental conditions.
Experimental validation is conducted in a semi-anechoic chamber, with the solar inverter placed on the floor and fans operating at 9,000 rpm. Twenty microphones capture SPL data, which are averaged and compared to simulation results. Table 2 summarizes the comparison of A-weighted SPL at key frequencies, demonstrating good agreement for tonal components but discrepancies in broadband noise due to unmodeled sources like vibration-induced radiation.
| Frequency (Hz) | Simulated SPL | Experimental SPL | Difference |
|---|---|---|---|
| 1,050 (Fundamental) | 45.2 | 46.8 | -1.6 |
| 2,100 (2nd Harmonic) | 42.7 | 43.5 | -0.8 |
| 3,150 (3rd Harmonic) | 39.4 | 40.1 | -0.7 |
| Broadband (Avg.) | 38.1 | 35.3 | +2.8 |
The overall A-weighted sound power level for the initial solar inverter design is measured at 69.16 dB(A). To reduce noise, we investigate the impact of the intake grille, which disturbs inflow uniformity and exacerbates pressure pulsations. By removing the grille structure, we observe a more uniform pressure distribution on the fan blades, as shown in Table 3, which compares pressure fluctuations before and after modification.
| Parameter | With Grille | Without Grille | Change |
|---|---|---|---|
| Max Negative Pressure (Pa) | -220 | -180 | +40 |
| Pressure Fluctuation Range (Pa) | 340 | 290 | -50 |
| Blade Surface SPL Max (dB) | 92 | 90 | -2 |
| Outlet Mass Flow (kg/s) | 0.0462 | 0.0465 | +0.0003 |
The reduction in pressure fluctuations leads to a decrease in radiated noise, as predicted by the acoustic simulation. The modified solar inverter exhibits lower SPL across most frequency bands, particularly at harmonics. The overall sound power level drops to 67.21 dB(A), a reduction of 1.95 dB(A), without compromising cooling performance, as the mass flow rate remains stable. This highlights the effectiveness of simple design changes in optimizing solar inverter acoustics.
Further analysis involves parametric studies on fan geometry and operating conditions. For instance, the blade passing frequency (BPF) and its harmonics are critical noise contributors in solar inverters. The BPF is given by:
$$f_{\text{BPF}} = \frac{N \times \text{RPM}}{60}$$
where \(N\) is the number of blades. For our solar inverter fans with 7 blades at 9,000 rpm, \(f_{\text{BPF}} = 1,050\) Hz. Higher harmonics can be expressed as \(f_n = n \times f_{\text{BPF}}\) for \(n = 1, 2, 3, \ldots\). Noise reduction strategies often target these frequencies through blade shaping or speed control.
We also explore the role of turbulence intensity in noise generation. The turbulent kinetic energy \(k\) and its dissipation rate \(\epsilon\) influence broadband noise, which can be estimated using empirical models. For solar inverters, minimizing turbulence in the inflow and outflow regions is crucial. The sound pressure level due to turbulent flow can be related to flow parameters by:
$$L_p \approx 10 \log_{10}\left( \frac{\rho^2 U^8 L^2}{c_0^4 r^2} \right) + K$$
where \(U\) is flow velocity, \(L\) is characteristic length, \(r\) is distance, and \(K\) is a constant. This formula emphasizes the sensitivity of noise to velocity, justifying lower fan speeds in solar inverters where possible.
In addition to aerodynamic noise, structural vibrations of the inverter housing can contribute to overall noise levels. However, this study focuses solely on fluid-borne noise sources. Future work could integrate fluid-structure interaction analyses for a comprehensive assessment of solar inverter noise.
To summarize, our hybrid CFD-CAA approach provides a reliable tool for predicting and optimizing aerodynamic noise in solar inverter cooling systems. Key findings include the dominant role of blade pressure pulsations, the significant impact of intake structures on flow uniformity, and the potential for noise reduction through design modifications. Solar inverters are essential for renewable energy adoption, and their acoustic performance must be prioritized to ensure user satisfaction and environmental compliance. We recommend that manufacturers consider streamlined intake designs and fan selection criteria during the development of solar inverters.
For further improvement, advanced turbulence models like Large Eddy Simulation (LES) could be employed for more accurate flow resolution, though at higher computational cost. Additionally, active noise control techniques or acoustic absorbers could be integrated into solar inverter housings. The continuous evolution of solar inverter technology will benefit from multidisciplinary approaches that balance thermal management, efficiency, and acoustics.
In conclusion, this research demonstrates the feasibility of using numerical simulations to address aerodynamic noise challenges in solar inverters. By incorporating real-world geometries and validating results experimentally, we offer practical insights for engineers designing quieter cooling systems. As the demand for solar inverters grows globally, such efforts will contribute to sustainable and noise-conscious energy solutions.