Computational analysis of flow-induced noise in shell-and-tube heat exchanger: Shell-side excitation and tube vibration dominate

Abstract

In heat exchangers, cold and hot fluid flow during heat exchange generates noise. This noise propagates through connected piping systems and considerably affects the acoustic stealth performance of vessels. With increasing flow rates and the trend toward larger-scale heat exchangers, noise issues have become more pronounced, highlighting the urgent need for effective noise control strategies. However, the underlying excitation mechanisms, noise sources, and radiation characteristics of flow-induced noise in liquid-cooled heat exchangers remain poorly understood. This gap renders vibration and noise control efforts ineffective and insufficient to support the low-noise design of heat exchangers. To address this gap, herein, we performed numerical analysis using computational fluid dynamics and acoustic analogy theory. Noise levels were calculated under both shell-side and tube-side flow conditions. A comparative analysis revealed that heat exchanger noise is primarily driven by shell-side flow excitation, specifically, the sound pressure level of noise induced by shell-side flow is approximately 26 dB higher than that induced by tube-side flow. Furthermore, under shell-side excitation, the radiated noise at the tube-side inlet and outlet is about 5 dB higher than that at the shell-side ports. Considering that the tube-side passages are directly connected to the vessel’s seawater piping system, this finding underscores the critical necessity of investigating and mitigating heat exchanger noise. Further, the contributions of different structural panels within the heat exchanger were analyzed. Color-mapped diagrams depicted the noise contributions of individual panels. Based on vector projection, a normalized panel contribution evaluation method was established to enable quantitative comparisons of the contributions of structural panels across different frequencies. The results indicated that acoustic radiation generated by heat exchanger tube vibrations constitutes the dominant noise source. Finally, experimental measurements were conducted to verify the accuracy of the numerical calculations. Overall, this study’s findings clarify the mechanism of flow-induced noise generation in heat exchangers and guide low-noise heat exchanger design, thereby supporting improvements in acoustic stealth performance.

Keywords

heat exchanger flow-induced noise generation mechanism computational analysis noise control

1. Introduction

Heat exchangers are widely used in power generation, chemical industry, HVAC, automotive, and marine engineering. Their thermal management role is critical to system efficiency and operational safety. Among the various types, the shell-and-tube heat exchanger is the most commonly used. As flow rates increase and exchanger sizes expand, flow-induced vibration and noise have become prominent concerns across these industries. In particular, for shipboard seawater cooling systems, noise generated by heat exchangers can propagate through connected piping and significantly degrade the acoustic stealth performance of vessels. Consequently, investigating flow-induced vibration and noise in heat exchangers is essential.

Accordingly, numerous studies have examined flow-induced vibrations in heat exchangers. For instance, Goyder (2002) and Thiago et al. (2017) investigated flow-induced vibrations in the tube bundles of heat exchangers and proposed control strategies, including increased support strength. Meanwhile, Khushnood and Nizam (2017) experimentally assessed tube bundle vibrations induced by shell-side cross-flow and compared vibration amplitudes at different flow velocities. Notably, beyond these flow velocity effects, interactions between adjacent tubes also influence vibration behavior. Accordingly, Liu et al. (2018) employed the finite element method to investigate vortex-induced vibration in a shell-and-tube heat exchanger containing two elastic tubes, focusing on tube interaction effects. Further, Hu et al. (2018) conducted wind tunnel experiments on vortex-induced vibrations in a staggered dual-tube configuration subjected to lateral flow and analyzed the corresponding vibration responses at different wind speeds. Xu et al. (2019a) investigated flow-induced vibrations in elastically mounted tandem tube bundles. At specific pitch-to-diameter ratios, the vibration behavior differed considerably from that of a single tube bundle, reflecting interaction effects between the two bundles. Further, Shaaban and Mohany (2018) examined vibration characteristics in tube bundles with unequal spacing under transverse flow excitation and compared their flow fields with those of equally spaced bundles to evaluate their effects on flow distribution. In another study, Chen et al. (2020) observed that excitation force, vibration amplitude, and vibration frequency depend on flow velocity and the spacing ratio of the tube bundle array. Beyond spacing effects, Lai et al. (2021) examined flow-induced vibrations in rotated triangular tube bundles and demonstrated that bundle arrangement also influences vibration behavior. To predict flow-induced vibrations in heat exchangers, Khulief et al. (2009) established a model for such vibrations in tube bundles, thereby enabling vibration calculations. Miu et al. (2018) and Ji et al. (2016) analyzed flow-induced vibrations in heat exchanger tubes using a fluid–structure interaction approach. Shahzer et al. (2022) numerically investigated the vortex-induced vibrations and flow-induced rotation of a horizontally placed elliptic cylinder. They found that the rotational degree of freedom amplified the transverse vibration amplitude by approximately 30% and widened the lock-in range compared with one or two degree of freedom systems. Their work was motivated by heat exchanger design, highlighting the potential of non-circular tube shapes in thermal applications.

Overall, the above studies predominantly focused on flow-induced vibrations in heat exchanger tube bundles. This emphasis reflects concerns regarding fatigue damage and equipment service life, which, although critical, are generally detectable and manageable through regular maintenance and proper operation. However, flow-induced noise is also present and is particularly important in acoustic stealth research. Such noise primarily affects tactical and technical performance metrics. Although it does not affect normal operation, it degrades vessel acoustic stealth and consequently influences survivability and combat effectiveness. Moreover, compared with flow-induced vibration, the mechanisms of flow-induced noise are more complex, making its quantitative evaluation and detection more challenging.

Given these issues, some researchers have examined acoustic issues in heat exchangers. For instance, Surendran et al. (2018, 2022) investigated the acoustic scattering behavior of heat exchanger tube rows in cross-flow. Cao et al. (2017, 2022) analyzed sound propagation in heat exchangers using cooperative field theory. Further, Allam (2015) established an acoustic model of heat exchangers using a two-port matrix and calculated sound transmission loss considering environmental gradient effects. Jiang et al. (2016, 2017a, 2017b, 2021), Xu et al. (2019b) and Hao et al. (2019) treated periodic tube bundle arrays as phononic crystal structures and investigated their attenuation effects on sound propagation. Czwielong et al. (2021) assessed the effects of densely packed tube bundles in heat exchangers on sound propagation and experimentally verified the acoustic attenuation performance of heat exchanger tube bundles in the Bragg frequency domain. Overall, these studies primarily examined acoustic transmission in heat exchangers rather than flow-induced noise generation. To address this, Baird (1954) first reported the flow-induced noise in heat exchangers. Subsequently, Blevins and Bressler (1987, 1993) experimentally investigated acoustic resonance in heat exchanger tube bundles, demonstrating that intense resonance occurs when the vortex shedding frequency coincides with transverse acoustic modes. Ziada et al. (1989) further revealed the mode competition mechanism that explains why some tube patterns suppress resonance despite frequency coincidence. Hamakawa et al. (2008) constructed a scaled model of an actual boiler plant and demonstrated that in-line tube banks with a small pitch ratio (L/D = 1.33) can excite higher-order acoustic resonances, even though natural vortex shedding does not directly match the resonance frequencies. Their results complement the observations of Blevins and Bressler (1987, 1993) and Ziada et al. (1989) and highlight the importance of considering multiple acoustic modes when assessing resonance risks in practical heat exchangers. Ding et al. (2022) investigated tube-type flue gas heat exchangers in coal-fired power plants and found that acoustic resonance induced by vortex shedding or turbulent flutter generates noise. Similarly, Hong et al. (2020) examined the frequency-locking mechanism of acoustic resonance, providing insights into internal noise generation in heat exchangers.

Regarding noise control techniques, Yamashita et al. (2014) investigated the mechanism of tonal noise generation from a circular cylinder with a spiral fin. Alziadeh and Mohany (2019) suppressed flow-excited acoustic resonance in heat exchanger tube bundles by employing non-uniform finned cylinders. Furthermore, they revealed the underlying mechanism of flow-induced acoustic resonance for finned cylinders (Alziadeh and Mohany, 2023), demonstrating that the fins and crimps suppress acoustic radiation efficiency by confining the acoustic particle velocity to the fin base region. Fiorentin et al. (2017) conducted a case study on an industrial cross-flow heat exchanger, identifying vortex-shedding-excited acoustic resonance as the cause of excessive noise and vibration, and successfully suppressed it by installing acoustic baffles. Similarly, Eisinger and Sullivan (2003) employed acoustic baffles to suppress acoustic resonance and vibration.

However, these investigations predominantly considered flow-induced noise in gas-medium heat exchangers. Meanwhile, in water-medium heat exchangers, the high velocity of sound and long wavelength of acoustic waves in water complicate the realization of internal acoustic resonance, leading to noise-generation mechanisms that differ from those observed in gas-medium systems. For such water-medium heat exchangers, Han et al. (2023) numerically examined radiation noise from shell-and-tube heat exchangers under lateral turbulent excitation. Wang et al. (2023) further computed the flow-induced noise characteristics of shell-and-tube heat exchangers using large eddy simulation and analyzed the effects of tube geometry, tube spacing, external flow velocity, and tube length on noise generation. Nevertheless, these studies did not provide a detailed analysis of the primary noise sources or the radiation patterns at the tube-side and shell-side inlets and outlets.

In summary, existing research on flow-induced noise in heat exchangers remains incomplete and does not clearly elucidate the underlying mechanisms. This limitation hinders effective acoustic assessment and the development of low-noise heat exchanger designs. To bridge this gap, the present study investigates the generation of flow-induced noise in heat exchangers. By integrating numerical simulations and experimental methods, it analyzes the mechanisms of flow-induced noise generation. Beyond identifying the fundamental generation mechanisms, we introduce a novel normalized contribution evaluation method that enables a quantitative, frequency-dependent assessment of individual structural panels to the overall radiated noise, thereby providing critical insights for the low-noise design of heat exchangers.

2. Numerical simulation of flow-induced noise

Here, computational fluid dynamics (CFD) and acoustic analogy theory were applied to analyze flow-induced noise in heat exchangers, following the computational workflow illustrated in Figure 1. First, CFD simulations captured the flow characteristics within the internal flow domain of the heat exchanger. The resulting pulsating pressures extracted from the flow field served as input for flow-induced noise calculations. These time-domain pulsating pressure data were saved in CFD general notation system format and subsequently imported into Virtual Lab software. The data were then converted into the frequency domain using fast Fourier transform. Because the software integrates a structural solver, unified structural and acoustic models could be directly established to solve acoustic–vibration coupling problems and subsequently derive noise characteristics.

Figure 1.

Computational workflow for calculating flow-induced noise in heat exchangers.

2.1. Flow-field computation

The complete geometrical data of the heat exchanger model are shown in Figure 2. The computational model comprises three parts: the solid structure, the tube-side fluid domain, and the shell-side fluid domain. The solid structure is made of steel, and both fluid domains are filled with water. The material properties are summarized in Table 1. The overall length of the heat exchanger is 1270 mm, and its outer diameter is 230 mm. Both the tube-side and shell-side pipe connections have a nominal diameter of DN80. Inside the shell, 35 tubes are arranged in a square pattern. Each tube has an outer diameter of 20 mm and a wall thickness of 1 mm. The tube pitch is 26 mm, yielding a pitch-to-diameter ratio of 1.3. Three baffles made of poly (vinyl chloride) (PVC) are installed inside the shell to reduce the flow-induced noise generated by the baffles themselves (Ciappi et al., 2012; Prajapati et al., 2023; Zuo and Zhang, 2025), thereby avoiding interference with the analysis.

Figure 2.

Schematic of the heat exchanger geometry and computational domains.

Table 1.

Material properties used in the numerical simulation.

Water			Steel
Density (kg/m³)	Viscosity (kg·s/m)	Sound velocity (m/s)	Density (kg/m³)	Young’s modulus (Pa)	Poisson’s ratio
1000	$1.003 \times 10^{- 3}$	1500	7850	$2 \times 10^{11}$	0.3

Mesh generation was performed using the Fluent mesh module with three refinement levels: coarse, fine, and ultra-fine. Numerical simulations were subsequently carried out using each mesh type. To verify mesh independence, the pressure difference P between the shell-side inlet and outlet, y+ value on the tube bundle surface, and tube bundle pressure were compared. The corresponding results are presented in Table 2. Evidently, as the number of grid cells increases from 12.76 million to 15.21 million, parameter variations remain minimal, and the calculated values for all parameters remain essentially consistent. Accordingly, the fine mesh was selected as the final computational mesh for the flow field.

Table 2.

Grid-independence verification.

Mesh type	Coarse mesh	Fine mesh	Ultra-fine mesh
Shell-side inlet/outlet pressure differential (Pa)	30856.54	31599.62	31784.81
Tube bundle y+ (average)	2.53	1.38	0.66
Average bundle pressure (Pa)	19136.19	19093.45	19514.77
Number of grids	10.82 million	12.76 million	15.21 million

Flow-field simulations were performed using Fluent software. The meshing of the tube-side and shell-side fluid domains was carried out according to the settings listed in Table 3. For the CFD simulations, a velocity inlet boundary condition was applied with an inlet velocity of 3 m/s, directed normal to the boundary. A pressure outlet boundary condition was used at the outlet with a gauge pressure of 0 Pa. All walls were set as no-slip stationary walls. The Fluent solver was configured according to Table 4. First, steady-state calculations were conducted. After the residuals dropped below

1 \times 10^{- 4}

and the flow field became reasonably developed, unsteady simulations were performed based on the steady-state results to compute the heat exchanger flow field and extract wall dipole acoustic sources. The time step for the transient calculations was set to

1 \times 10^{- 4}

Table 3.

Mesh generation settings for the computational domains.

Physics preference	Element order	Element size	Adaptive sizing	Inflation layer
Physics preference	Element order	Element size	Adaptive sizing	First layer height	Maximum layers	Growth rate
CFD	Linear	3 × 10^-3m	Yes	5 × 10^-5m	16	1.2

Table 4.

Numerical setup for the Fluent solver.

Settings		Steady	Transient
Solver type		Pressure-based	Pressure-based
Viscous model		SST k-omega	Smagorinsky-Lilly LES
Pressure-velocity coupling		Simple	PISO
Spatial discretization	Gradient	Least squares cell based	Least squares cell based
	Pressure	Standard	PRESTO
	Momentum	First order upwind	Bounded central differencing
	Turbulent kinetic energy	First order upwind	/
	Specific dissipation rate	First order upwind	/
Transient formulation		/	Bounded second order implicit

Figure 3 compares the computed velocity field contours and the pulsating pressure contours of the tube-side and shell-side domains. It is evident that, compared with the tube-side flow, the shell-side flow is more unstable and exhibits locally higher velocities, reaching up to 6 m/s. Furthermore, the amplitude of the pulsating pressure in the shell-side flow field (maximum value: 48877.7 Pa) is significantly higher than that in the tube-side flow field (maximum value: 18086.6 Pa).

Figure 3.

Comparison of velocity and pulsating pressure fields. (a) Tube-side velocity. (b) Shell-side velocity. (c) Tube-side pulsating pressure. (d) Shell-side pulsating pressure.

2.2. Acoustic calculations

Acoustic calculations were performed using the transient flow-field results. Figure 4 presents the noise levels at the tube-side and shell-side inlet and outlet of the heat exchanger under shell-side flow conditions. Overall, the acoustic energy is concentrated in the low-frequency range, while the sound pressure level gradually decreases as frequency increases. Furthermore, the noise levels at the tube-side inlet and outlet are approximately 5 dB higher than those at the shell-side inlet and outlet. Because the tube side connects directly to the cooling water circuit, its noise contributes more prominently to the acoustic radiation of that circuit.

Figure 4.

Sound pressure levels at the inlet and outlet ports under shell-side flow conditions. (a) Tube-side inlet/outlet noise. (b) Shell-side inlet/outlet noise.

To evaluate the contributions of structural vibrations at various locations within the heat exchanger to noise generation, the exchanger was divided into multiple structural panels, as illustrated in Figure 5. Under shell-side flow conditions, acoustic radiation induced by the structural vibration of each panel was computed based on the pulsating pressure acting on the wall surface obtained from the CFD simulations. Subsequently, the noise contributions from the different panels were compared to identify the primary acoustic radiation sources.

Figure 5.

Schematic of the heat exchanger structural panels.

Figure 6 presents a color-mapped comparison of the noise contributions of the structural panels of the heat exchanger under shell-side flow conditions. Here, three structural panel types are considered: tube, tubesheet, and shell. In the figures, color intensity represents the noise contribution level, while “Total” denotes the overall noise contribution. Notably, the color distribution of the tube panel closely matches that of the Total distribution. This correspondence indicates that the tubes dominate noise generation at both the shell-side and tube-side inlet and outlet. Meanwhile, the color patterns of the other panels exhibit weaker agreement with the Total distribution. In some frequency bands, individual panels even appear darker than the Total color. This finding suggests that, within those frequency ranges, these panels suppress noise generation because their vibration occurs in an opposite phase.

Figure 6.

Noise contributions from different structural panels measured at inlet and outlet under shell-side flow conditions. (a) Tube-side inlet. (b) Tube-side outlet. (c) Shell-side inlet. (d) Shell-side outlet.

To quantitatively compare noise contributions from different structural panels, a vector projection method was employed to determine the contribution magnitude of each panel. Specifically, the contribution vector of each panel was projected onto the direction of the total vector. When the projection direction aligned with the total vector, the contribution was deemed positive; otherwise, it was considered negative. The projection length therefore represented the effective contribution magnitude. Overall, the projection length of a structural panel onto the total vector was calculated using the following expression:

C_{p r o j} (i, f) = \frac{P_{p a n e l} (i, f) \cdot P_{t o t a l} (f)}{| P_{t o t a l} (f) |},

(1)

where C_proj (i,f) represents the projection length of structural panel i at frequency f. P_panel (i,f) denotes the complex sound pressure of panel i at frequency f, while P_total(f) represents the total noise sound pressure at the response point at frequency f. To enable contribution comparison across different frequencies and highlight the relative importance of each panel, parameter normalization was subsequently performed. For each frequency f, the maximum absolute value of the projected contribution among all panels was first identified. The projected contributions of all panels at that frequency were then divided by this maximum value, yielding normalized panel contributions within the range [−1, 1]. The normalization procedure is presented in equation (2):

C_{n o r m} (i, f) = \frac{C_{p r o j} (i, f)}{\max_{j} | C_{p r o j} (j, f) |} .

(2)

Using the normalized panel contribution method described above, the contributions of each structural panel were analyzed under shell-side flow conditions. As depicted in Figures 7 and 8, the tube structural panel consistently exhibits higher normalized contribution values than the other structural panels. Furthermore, its normalized contributions remain positive at all noise observation points. This pattern indicates that the heat transfer tube bundle dominates noise generation at both the tube-side and shell-side inlet and outlet, particularly near 150 Hz where the acoustic energy is concentrated. In comparison, the tubesheet contributes only weakly, with a noticeable contribution appearing near 50 Hz. Moreover, the tubesheet contributions at the two observation locations exhibit opposite phases. In contrast, the shell structural panel contributes negligibly and is largely overshadowed by the other structural panels.

Figure 7.

Normalized noise contributions from structural panels measured at tube-side inlet and outlet under shell-side flow conditions. (a) Tube-side inlet. (b) Tube-side outlet.

Figure 8.

Normalized noise contributions from structural panels measured at shell-side inlet and outlet under shell-side flow conditions. (a) Shell-side inlet. (b) Shell-side outlet.

Figure 9 presents the sound pressure levels at the heat exchanger inlet and outlet under tube-side flow conditions. Notably, the tube-side inlet and outlet exhibit higher sound pressure levels, exceeding those at the shell-side inlet and outlet by more than 1 dB. This result indicates that the tube-side inlet and outlet generate the highest noise levels in the heat exchanger regardless of whether the shell-side or tube-side fluid is flowing. Because these locations connect directly to the seawater pipeline, this observation further emphasizes the importance of investigating noise within heat exchangers.

Figure 9.

Sound pressure levels at each inlet and outlet under tube-side flow conditions. (a) Tube-side inlet/outlet noise. (b) Shell-side inlet/outlet noise.

Based on the above analysis, a panel contribution analysis was conducted to compute the noise generated by vibrations in each structural panel and compare their respective contributions. As depicted in Figure 10, the color distribution of the tube panel closely matches the Total distribution, consistent with the shell-side flow condition. This correspondence indicates that vibrations of the heat exchange tubes dominate noise generation at all four inlet and outlet locations under tube-side flow.

Figure 10.

Noise contributions from different structural panels measured at inlet and outlet under tube-side flow conditions. (a) Tube-side inlet. (b) Tube-side outlet. (c) Shell-side inlet. (d) Shell-side outlet.

However, unlike the shell-side flow condition, the tubesheet structural panel exhibits notable contributions within the frequency ranges of 230 Hz – 380 Hz and 700 Hz – 800 Hz during tube-side fluid flow, as illustrated in Figures 11 and 12. This behavior results from strong flow impacts at the tubesheet during tube-side fluid flow.

Figure 11.

Normalized noise contributions from structural panels measured at tube-side inlet and outlet under tube-side flow conditions. (a) Tube-side inlet. (b) Tube-side outlet.

Figure 12.

Normalized noise contributions from structural panels measured at shell-side inlet and outlet under tube-side flow conditions. (a) Shell-side inlet. (b) Shell-side outlet.

The sound pressure levels at the tube-side and shell-side outlets were compared under both shell-side and tube-side flow conditions. As depicted in Figure 13, shell-side flow generates considerably higher sound pressure levels than tube-side flow. Quantitatively, the tube-side outlet exhibits a sound pressure increase of approximately 26 dB, whereas the shell-side outlet exhibits a corresponding increase of approximately 22 dB. These results indicate that shell-side fluid flow dominates noise excitation in the heat exchanger.

Figure 13.

Comparison of noise under shell-side and tube-side flow conditions. (a) Tube-side outlet noise. (b) Shell-side outlet noise.

3. Experimental test

The accuracy of the computational results was experimentally validated. Figure 14 illustrates the corresponding test setup. As discussed earlier, shell-side fluid flow primarily generates noise in heat exchangers. Furthermore, the tube-side inlet and outlet exhibit higher noise levels under both shell-side and tube-side flow conditions. Therefore, the validation focused on noise levels measured at the tube-side inlet and outlet during shell-side flow conditions. To eliminate the influence of tube-side flow on hydroacoustic signal measurements, the tube-side flow was stopped, allowing only shell-side fluid to flow at a regulated velocity of 3 m/s.

Figure 14.

Test setup. (a) Structural schematic of the heat exchanger. (b) Site photo of the experiment.

The sound pressure was measured using a RESON TC4013 hydrophone, which has a frequency response of 1 Hz to 170 kHz and a sensitivity of −211 dB ± 3 dB (ref $1 V / μ Pa$ ). The hydrophone was positioned at the tube-side inlet and outlet, flush with the pipe wall. The signals were acquired using a Brüel & Kjær 2692 conditioning amplifier and a Brüel & Kjær 3050-A-060 input module, with a Nyquist frequency of 12.8 kHz, for 30 s per measurement. The sound pressure level (SPL) in water was calculated as:

S P L = 20 \log_{10} (\frac{p_{r m s}}{p_{r e f}})

(3)

where

p_{r m s}

is the root-mean-square sound pressure, and

p_{r e f} = 1 μ Pa

is the standard reference pressure for underwater acoustics. Frequency-domain analysis was performed by applying a Fast Fourier Transform (FFT) to the time-domain pressure signals. A Hanning window with 66.67% overlap was used to reduce spectral leakage. The frequency resolution was

Δ f = 2 Hz

. Because the noise energy is concentrated mainly in the low-frequency range and the high-frequency measurements exhibit a low signal-to-noise ratio, the 10 Hz-1 kHz frequency band was selected for analysis. To minimize measurement errors, the procedure was repeated three times, and the average value of the hydroacoustic signals was used.

Figure 15 compares the calculated and experimental noise levels at the tube-side inlet and outlet under shell-side flow conditions. At the tube-side inlet, the calculated overall sound pressure level (OASPL) is 155.7 dB, while the experimental value is 160.4 dB, yielding a difference of 4.7 dB. At the tube-side outlet, the calculated OASPL is 155.2 dB and the experimental value is 160.2 dB, giving a difference of 5.0 dB. Several factors may contribute to the discrepancies between the computed and experimental noise levels observed in Figure 15. First, although vibration and noise control measures were applied to the shell-side pump, they could not be completely eliminated. Residual pump noise transmitted through the shell-side fluid and structure inevitably coupled into the tube-side circuit, thereby elevating the measured noise levels at the tube-side inlet and outlet. Second, acoustic reflections occurred in the piping system connected to the tube side. These reflections can interfere with the direct acoustic signal and produce constructive or destructive effects at specific frequencies, leading to deviations between the measured and predicted spectra. Third, suboptimal welding quality has resulted in partial obstructions at the flange connections, which affect sound transmission and consequently introduce uncertainties into the experimental results.

Figure 15.

Comparison of experimental and computational noise levels. (a) Tube-side inlet. (b) Tube-side outlet.

Despite these discrepancies, the overall spectral trends show good agreement between the computation and the experiment, and the differences in the total sound pressure level are within 5.0 dB. Therefore, the comparison confirms the accuracy of the computational results for engineering purposes.

4. Conclusion

In this study, we performed a computational analysis of flow-induced noise in heat exchangers. Consequently, the primary noise excitation mechanisms were identified, and the dominant noise-contributing components were clarified. The following main conclusions were derived:

(1) Internal fluid flow within the heat exchanger generates substantial noise at the tube-side inlet and outlet connected to the seawater pipeline. The resulting noise propagates along the piping and degrades acoustic stealth performance, emphasizing the importance of investigating flow-induced noise in heat exchangers.

(2) Compared with tube-side flow, shell-side flow produces substantially higher sound pressure levels, with increases of approximately 26 dB at the tube-side outlet and 22 dB at the shell-side outlet. These results indicate that shell-side fluid flow constitutes the primary noise excitation source in the heat exchanger.

(3) A normalized panel contribution calculation method based on vector projection was established to compare the noise contributions of different structural panels within the heat exchanger at the inlet and outlet. This analysis revealed that heat exchanger noise is primarily generated by tube vibrations.

Overall, these findings provide valuable guidance for the acoustic evaluation and low-noise design of heat exchangers. Nevertheless, the baffles are made of PVC, which produce negligible flow-induced noise; therefore, they were excluded from the panel contribution analysis. Future work should incorporate baffle-induced noise, particularly for metallic baffles. Moreover, the absence of a systematic parametric analysis restricts the generalizability of the present results to heat exchangers with substantially different structural configurations. Experimentally, future test setups should be better designed to isolate shell-side noise sources (e.g., pump noise), and an anechoic termination should be used on the tube side to approximate a non-reflecting boundary.

Footnotes

ORCID iD

Wenhua Chen

CRediT authorship contribution statement

Wenhua Chen: Investigation, Writing—original draft, Review and editing, Validation.

Xueguang Liu: Formal analysis, Supervision.

Weixi Huang: Methodology, Visualization.

Yun Liang: Data curation.

Yudong Sun: Conceptualization.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the National Key Laboratory on Ship Vibration and Noise under Grant No. WDZC70202020102.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

Data will be made available on request.*

References

Allam

(2015) Effect of ambient gradients on sound transmission in narrow permeable rectangular pipes with application to heat exchangers. Advances in Powertrains and Automotives 1(1): 24–33. https://doi.org/10.12691/apa-1-1-3

Alziadeh

Mohany

(2019) Passive noise control technique for suppressing acoustic resonance excitation of spirally finned cylinders in cross-flow. Experimental Thermal and Fluid Science 102: 38–51. https://doi.org/10.1016/j.expthermflusci.2018.10.029

Alziadeh

Mohany

(2023) Flow structure and aerodynamic forces of finned cylinders during flow-induced acoustic resonance. Journal of Fluids and Structures 119: 103887. https://doi.org/10.1016/j.jfluidstructs.2023.103887

Baird

(1954) Pulsation-induced vibration in utility steam generation units. Combustion 25(10): 38–44.

Blevins

Bressler

(1987) Acoustic resonance in heat exchanger tube bundles-part I: physical nature of the phenomenon. Journal of Pressure Vessel Technology 109(3): 275–281. https://doi.org/10.1115/1.3264863

Blevins

Bressler

(1993) Experiments on acoustic resonance in heat exchanger tube bundles. Journal of Sound and Vibration 164(3): 503–533. https://doi.org/10.1006/jsvi.1993.1231

Cao

Lin

, et al. (2017) Investigation on the flow noise propagation mechanism in pipelines of shell-and-tube heat exchangers based on synergy principle of flow and sound fields. Applied Thermal Engineering 122: 339–349. https://doi.org/10.1016/j.applthermaleng.2017.04.057

Cao

, et al. (2022) Investigation on the Acoustic energy transfer process in expanded pipe of heat exchangers. Heat Transfer Engineering 43(8–10): 679–693. https://doi.org/10.1080/01457632.2021.1905299

Chen

, et al. (2020) Flow-induced vibrations of two side-by-side circular cylinders at low reynolds numbers. Physics of Fluids 32(2): 1–18. https://doi.org/10.1063/1.5129013

10.

Ciappi

Magionesi

Rosa

, et al. (2012) Analysis of the scaling laws for the turbulence driven panel responses. Journal of Fluids and Structures 32: 90–103. https://doi.org/10.1016/j.jfluidstructs.2011.11.003

11.

Czwielong

Hruska

Bednarik

, et al. (2021) On the acoustic effects of sonic crystals in heat exchanger arrangements. Applied Acoustics 182: 1–17. https://doi.org/10.1016/j.apacoust.2021.108253

12.

Ding

, et al. (2022) Analysis on fluid-induced vibration of tubular gas-gas heat exchanger for ultra-supercritical 660 MW coal-fired generating unit. Thermal Power Generation 51(12): 99–105. https://doi.org/10.19666/j.rlfd.202207146

13.

Eisinger

Sullivan

(2003) Suppression of acoustic waves in steam generator and heat exchanger tube banks. Journal of Pressure Vessel Technology 125: 221–227. https://doi.org/10.1115/1.1565079

14.

Fiorentin

Mikowski

Silva

, et al. (2017) Noise and vibration analysis of a heat exchanger: a case study. International Journal of Acoustics and Vibration 22(2): 270–275. https://doi.org/10.20855/ijav.2017.22.2473

15.

Goyder

HGD

(2002) Flow-induced vibration in heat exchangers. Chemical Engineering Research and Design 80(3): 226–232. https://doi.org/10.1205/026387602753581971

16.

Hamakawa

Matsue

Nishida

, et al. (2008) Acoustic resonance and vortex shedding from tube banks of boiler plant. Journal of Fluid Science and Technology 3(6): 805–813. https://doi.org/10.1299/jfst.3.805

17.

Han

Cao

Liu

, et al. (2023) Numerical study on flow-induced vibration and radiated noise of tube bundle in a shell-and-tube heat exchanger. Inter-Noise and Noise-Con Congress and conference Proceedings 268(7): 1972–1982. https://doi.org/10.3397/in_2023_0295

18.

Hao

Jiang

Liu

, et al. (2019) Propagation characteristics of apparent sound velocity in a periodic tube array. Technical Acoustics 38(5): 29–30.

19.

Hong

Wang

Jing

, et al. (2020) Frequency lock-in mechanism in flow-induced acoustic resonance of a cylinder in a flow duct. Journal of Fluid Mechanics 884: A42. https://doi.org/10.1017/jfm.2019.966

20.

Wang

, et al. (2018) Wind tunnel experiment on vortex-induced vibration of staggered double circular tube. Chemical Engineering & Machinery 45(6): 675–679. https://doi.org/10.3969/j.issn.0254-6094.2018.06.005

21.

(2016) Numerical analysis on shell-side flow induced vibration responses of multi-row elastic tube bundles in heat exchangers. Journal of Vibration and Shock 35(20): 85–89. 136. https://doi.org/10.13465/j.cnki.jvs.2016.20.014

22.

Jiang

Liu

Kong

, et al. (2016) The sound transmission through tube arrays in power boilers based on phononic crystals theory. Applied Thermal Engineering 99: 1133–1140. https://doi.org/10.1016/j.applthermaleng.2016.01.137

23.

Jiang

Liu

, et al. (2017a) Acoustic radiation and cold point defect in boiler tube arrays based on phononic crystals theory. Applied Thermal Engineering 123: 746–752. https://doi.org/10.1016/j.applthermaleng.2017.05.103

24.

Jiang

Liu

, et al. (2017b) Transmission and radiation of acoustic oblique incident through tube arrays based on phononic crystals theory. Applied Acoustics 116: 117–126. https://doi.org/10.1016/j.apacoust.2016.09.020

25.

Jiang

Liu

Kong

, et al. (2021) Study on acoustic time delay localization in boiler tube arrays considering effective sound velocity. Applied Acoustics 171: 1–8. https://doi.org/10.1016/j.apacoust.2020.107680

26.

Khulief

Ai-Kaabi

Said

, et al. (2009) Prediction of flow-induced vibrations in tubular heat exchangers—part I: numerical modeling. Journal of Pressure Vessel Technology 131(1): 011301. https://doi.org/10.1115/1.3006950

27.

Khushnood

Nizam

(2017) Experimental study on cross-flow induced vibrations in heat exchanger tube bundle. China Ocean Engineering 31(1): 91–97. https://doi.org/10.1007/s13344-017-0011-8

28.

Lai

Tan

Yang

, et al. (2021) Flow-induced vibration of tube bundles considering the effect of periodic fluid force in a rotated triangular tube array. Annals of Nuclear Energy 161: 108488. https://doi.org/10.1016/j.anucene.2021.108488

29.

Liu

Yue

, et al. (2018) Vortex-induced vibration characteristics of double elastic tubes in shell-and-tube heat exchanger. Chemical Engineering & Machinery 45(4): 495–500. 514. https://doi.org/10.3969/j.issn.0254-6094.2018.04.026

30.

Miu

Chen

, et al. (2018) Flow-induced vibration analysis of heat transfer tube based on two-way fluid-structure interaction. Nuclear Techniques 41(5): 1–7. https://doi.org/10.11889/j.0253-3219.2018.hjs.41.050602

31.

Prajapati

Anantharamu

Mahesh

(2023) Direct numerical simulation-based vibroacoustic response of plates excited by turbulent wall-pressure fluctuations. Journal of Fluid Mechanics 976(A2): 1–39. https://doi.org/10.1017/jfm.2023.870

32.

Shaaban

Mohany

(2018) Flow-induced vibration of three unevenly spaced in-line cylinders in cross-flow. Journal of Fluids and Structures 76: 367–383. https://doi.org/10.1016/j.jfluidstructs.2017.10.007

33.

Shahzer

Khan

Anwer

, et al. (2022) A comprehensive investigation of vortex-induced vibrations and flow-induced rotation of an elliptic cylinder. Physics of Fluids 34: 033605. https://doi.org/10.1063/5.0079642

34.

Surendran

Heckl

Peerlings

, et al. (2018) Aeroacoustic response of an array of tubes with and without bias-flow. Journal of Sound and Vibration 434: 1–16. https://doi.org/10.1016/j.jsv.2018.07.022

35.

Surendran

Boakes

, et al. (2022) A low frequency model for the aeroacoustic scattering of cylindrical tube rows in cross-flow. Journal of Sound and Vibration 527: 116806. https://doi.org/10.1016/j.jsv.2022.116806

36.

Thiago

Alexandre

(2017) Noise and vibration analysis of a heat exchanger: a case study. International Journal of Acoustics and Vibration 22(2): 270–275. https://doi.org/10.20855/ijav.2017.22.2473

37.

Wang

Liu

Sha

, et al. (2023) Large eddy simulation-based analysis of the flow-induced noise characteristics of shell and tube heat exchanger. International Communications in Heat and Mass Transfer 148: 107030. https://doi.org/10.1016/j.icheatmasstransfer.2023.107030

38.

Jiang

Peng

, et al. (2019a) Experimental study of acoustically enhanced heat-transfer from copper sphere in a steady flow field. IOP Conference Series: Materials Science and Engineering 626: 1–8. https://doi.org/10.1088/1757-899X/626/1/012014

39.

Sun

, et al. (2019b) Flow-induced vibration of two elastically mounted tandem cylinders in cross-flow at subcritical reynolds numbers. Ocean Engineering 173: 375–387. https://doi.org/10.1016/j.oceaneng.2019.01.016

40.

Yamashita

Hayashi

Okumura

, et al. (2014) Mechanism of tonal noise generation from circular cylinder with spiral fin. Journal of Thermal Science 23(6): 552–557. https://doi.org/10.1007/s11630-014-0740-4

41.

Ziada

Oengören

Bühlmann

(1989) On acoustical resonance in tube arrays part I: experiments. Journal of Fluids and Structures 3(3): 293–314. https://doi.org/10.1016/S0889-9746(89)90083-2

42.

Zuo

Zhang

(2025) Analysis of flow-induced acoustic radiation characteristics of composite material command chamber. Journal of Wuhan University of Technology 49(6): 1252–1257. https://doi.org/10.3963/j.issn.2095-3844.2025.06.013