Abstract Pipeline solid particle erosion depends on numerous independent variables and fluid flow parameters, making measurements or predictions inherently uncertain. Estimating these uncertainties is challenging due to a limited understanding of variable interactions. Because erosion is a major risk in many industries, accurately estimating uncertainties is vital for both safety and economic viability. This study examines uncertainties in previously gathered erosion measurements under two-phase flow conditions with sand in pipe elbows. The database includes ultrasonic measurements of gas–liquid–sand cases with vertical upward flows through standard elbows. Particle sizes range from 20 to 300 μm, and pipe diameters are 0.0762 and 0.1016 m. Measurements span multiple two-phase flow regimes, including annular, churn, and annular-mist flows. Initial uncertainty bounds were estimated using analytical correlations, yielding upper bounds of 1.45–2.78 times the measurements and lower bounds of 0.18–0.78 times the measurements. Subsequently, machine learning (ML) algorithms, including random forests, LightGBM, XGBoost, and Gaussian processes, were trained on the erosion database. These algorithms used gas and liquid velocities, pipe diameters, and particle sizes as independent variables, along with the initially estimated bounds. Finally, Monte Carlo (MC) simulations examined how uncertainties in these parameters affect the overall uncertainty. The MC-based bounds refined the original estimates, providing an updated range of 1.5–4.00 times the measurements for the upper bound and 0.04–0.78 times for the lower bound. This integrated approach provides a more accurate representation of pipeline erosion uncertainty, improving flow assurance across multiple industries.

Abstract In this work, the Delayed eXtra-Large Eddy Simulation (Delayed eXtra-LES) model is assessed for the simulation of the flow through a centrifugal pump, consisting of impeller, diffuser, and volute, both at design and off-design conditions. The results of the simulations of the impeller and pump with low- and high-fidelity models are compared with experiments and other numerical results available in literature to demonstrate the effectiveness in modeling complex flow phenomena in centrifugal pumps with a good compromise between fidelity and computational cost of the Delayed eXtra-LES model. A critical aspect is represented by the choice of the filter width, as different filter widths lead to different results. Furthermore, the simulations with the high-fidelity model of the impeller and complete pump demonstrated the influence of the volute on the flow field. The steady “two-channel” stall phenomenon with an alteration of stall and nonstall channels appears at the off-design conditions, both without and with the volute. The tongue of the volute as well as different turbulence models, Sub-Grid Scale models, and filter widths lead to differences in the dimensions and location of vortices.

Abstract The simplifying assumption of uniform porosity in studies of wave interaction with porous submerged discs is a major limitation in practical scenarios. This work presents a novel analysis of the heave motion of a circular disc with radially varying porosity, a configuration not addressed previously. Using an integral equation formulation, we evaluate the added mass and damping coefficients. The results reveal some deviations from uniform cases, highlighting the critical influence of spatial porosity gradients. These findings provide valuable insights into the hydrodynamic behaviour of porous structures with spatially varying properties, enabling more accurate modelling and optimized design of circular porous structures.

Abstract A method is proposed for reconstructing spanwise cross-sectional data of a cavitating turbulent flow field around a 2D hydrofoil to a 3D flow field using supervised machine learning (ML). One of the variables of interest was the cross-sectional pseudomeasurement data of particle image velocimetry. Therefore, the spanwise component w of velocity was not obtained; that is, it cannot be input into neural networks (NN). To compensate for this, we introduced a mass conservation error into the loss function of ML. The computational fluid dynamics (CFD) data for the training were obtained using large-eddy simulation with a homogeneous fluid model for cavitation. The super-resolution convolutional neural network (SRCNN) was adopted for the supervised ML framework. As a result, under the attached sheet cavitation mode, which corresponds to a quasi-steady condition, prediction was relatively simple. In contrast, under sheet/cloud cavitation mode, which is a strongly unsteady condition, the learning curve exhibited oscillations, suggesting that learning is difficult. Physical variables with a uniform distribution in the homogeneous spanwise direction can be reconstructed with high accuracy even under unsteady conditions, whereas the rotational components of streamwise vortices, which are nonuniform in the spanwise direction and have high spatiotemporal frequencies, fluctuate even under quasi-steady conditions, making them extremely difficult to reconstruct. When the loss function is augmented with a mass conservation error, the accuracy of the w reconstruction is improved as the NN attempts to satisfy the conservation of mass.

Abstract The combination of labyrinth seal and fluid curtain seal can significantly reduce the leakage rate. A loss assessment method of fluid curtain-labyrinth seals in axial turbines with supercritical carbon dioxide as working medium is proposed, and the action mechanism of key parameters is systematically investigated in this paper. The results indicate that combing the loss assessment method with the simplified model calculations can effectively evaluate the loss characteristics of different fluid curtain-labyrinth sealing structures, saving assessment time greatly. With regards to the action mechanism of key parameters on seal performance, as jet angle and seal cavity height decrease, leakage flow rate and loss decrease. A ratio of the jet width to the seal cavity height of 0.1 will minimize leakage losses. The proposed method can be applicable to various types of seals for any working fluid, and the action mechanism can provide guidance for the design and optimization of fluid curtain-labyrinth seals.

Abstract The growing integration of renewable energy, especially wind power, in urban settings necessitates the optimization and performance enhancement of small-scale vertical axis wind turbines (VAWTs). The major challenges in VAWT performance enhancement are impediments to self-starting, lower performance, flow instability due to blade wake interactions, and dynamic stability at lower tip speed ratios (TSRs). This work examines two aspects: the numerical investigation of the influence of TSR on the aerodynamic performance of the two-bladed H-rotor, and the analysis of the deep dynamic stall phenomenon involving interaction between blade and shaft vortices. It is observed that fluctuations increase to 2.6 times for the deep stall condition compared to no stall. It is also observed that between no stall and deep stall conditions, there is a significant difference (∼83%) in the peak-to-peak amplitude of the lift coefficient, indicating deterioration in rotor performance as it enters a deep stall. In the downwind zone, the vorticity contour indicates blade–vortex interaction, affirming the reduction in moment and its reduced recovery. The results of Fast Fourier Transform indicate that the peak amplitude decreases by a factor of 1.6 in the downwind zone, signifying the development of stronger vortical structures in the upwind region.

Abstract The loss generation in the corner region of compressors is highly dependent on vortex dynamics. When the Reynolds number (Re) decreases to a critical value, the flow unsteadiness induced by multiscale vortex structures changes significantly, resulting in difficulties in loss quantification and control. In this study, a high-subsonic compressor cascade was taken as the research object, and the low-Re-effects on the flow unsteadiness and loss generation in the corner region were clarified. A new source of flow unsteadiness in the corner region, namely,, low-frequency oscillation (LFO), which is induced by the resonance among multiple vortex structures with different frequencies, is first identified. When Re decreases from 5.6 × 105 to 2.5 × 105, more energy is transferred to LFO through the resonance process. As such, the region of LFO is distinctly extended, which becomes the most important source of flow unsteadiness in the corner region at a low Re. Furthermore, the contributions of various flow unsteadiness on the generation of TKE are quantified via multiscale proper orthogonal decomposition (mPOD) analysis, with the results showing that LFO contributes the largest proportion to TKE at a low Re. Thus, the loss increases or performance deterioration caused by the low-Re-effect is attributed mainly to the enhancement of LFO. That is to, say, the top priority of loss reduction in the corner region at a low Re is to suppress LFO growth by weakening the multiscale vortex-induced resonance process.

Abstract A biharmonic stream function equation is solved analytically in axisymmetric toroidal coordinates for Stokes flow in a sessile droplet with a surface tension gradient and diffusive evaporation on the surface. Four different types of streamlines are defined and utilized to predict particle deposition either at the center or at the contact line of the droplet. The impact of UV light radius, droplet's contact angle, and humidity level on the flow pattern is investigated. When the relative humidity is high, the formation of Moffatt eddies is observed. However, the eddy size and the number of eddies decrease with an increase in UV light radius or droplet contact angle. As humidity decreases, the Moffatt eddies are restricted to a specific region, decrease in number and size, and ultimately disappear. Under lower humidity, the Marangoni flow near the contact line is dominated by capillary action. The toroidal streamlines are broken from outside and turned into transporting streamline to deposit more particles on the contact line. The region near the droplet center, where the surface tension gradient is significant, is always dominated by toroidal streamlines.

Abstract This work investigates the current wall-modeled large-eddy simulation (WMLES) capabilities of the open-source computational fluid dynamics solver OpenFOAM, which is used widely in academia and industry. This is achieved by a simulation campaign that covers both attached and smooth body separation cases. The campaign includes simulations using four different wall models and aims to investigate the sensitivity of the results to changes in numerical schemes, mesh resolution, and subgrid-scale modeling. The results demonstrate that two main factors largely determine OpenFOAM-based WMLES performance. These are the discretization of the convective term and wall modeling. For the former, the best performance in the attached case is achieved with low-dissipation numerical schemes, however, for the smooth body separation case, the picture is more complicated. For the latter, we find that both equilibrium and non-equilibrium wall models perform well in the attached case but that the non-equilibrium models significantly improve the prediction of smooth body separation. Still, the non-equilibrium wall model results do not show a uniform improvement over equilibrium models. This is explained by an inconsistent accounting of non-equilibrium physics in these models, i.e., including the pressure gradient term without also including the convective term. This highlights the potential for future performance improvements by using non-equilibrium wall models that consistently account for both the convective and pressure gradient terms.

Abstract There is a significant need to reduce the aerodynamic drag of ground vehicles. As a result, the active and passive methods of flow control around bluff bodies have been extensively studied to improve aerodynamic performance. In this study, the drag reduction by combining rear flaps and plasma actuators (PAs) is experimentally investigated. PAs are modern active flow control devices that leverage nonthermal plasma, whereas flaps are traditional passive flow control devices. Flaps and PAs are mounted on the square back of the truck model, and the effect of the flap mounting locations (top or side of the model), flap angle, and flap length are investigated at ReH=1.58×105. For both top and side flaps, the longer flaps result in greater drag reduction compared to shorter flaps in the absence of PA actuation. However, additional drag reduction with PA actuation is minimal for the longer flaps, and the flap angle at which drag reaches a local minimum remained unchanged. This is attributed to an increase in pressure drag on the flap as the flap angle increases when the flow is attached to the flap surface. In contrast, PA actuation with shorter flaps demonstrates drag reduction, due to the smaller surface area of the flaps. These results suggest that optimizing the flap angle and length is a promising strategy to maximize the aerodynamic benefits of PA actuation and reduce the drag on bluff bodies.