Particle Image Velocimetry Experiments Research Articles

Experimental study on airfoil flow separation control via an air-supplement plasma synthetic jet

An air-supplement plasma synthetic jet (PSJ) actuator increases the air supplemental volume in the recovery stage and improves the jet energy by attaching a check valve to the chamber of a conventional actuator. To explore the flow control effect and mechanism of the air-supplement actuator, via particle image velocimetry experiments in a low-speed wind tunnel, the flow field and boundary layer characteristics of a two-dimensional airfoil surface under different actuation states were compared for different attack angles and jet orifices. The experimental results show that, compared with the conventional actuation state, the jet energy of the air-supplement PSJ is higher and the indirect mixing effect of the counter-vortex sequence produced by the jet-mainstream interaction is stronger. Furthermore, the boundary layer mixing effect is better, which can further suppress flow separation and improve the critical flow separation attack angle. Moreover, increasing the jet momentum coefficient can enhance the flow control effect. The findings of this study could provide guidance for the flow control application of air-supplement PSJs.

Development and validation of non-invasive endothelial shear stress (ESS) estimations of coronary artery haemodynamics using computational modelling

Abstract Background Vascular remodelling is influenced by various biomechanical forces such as endothelial shear stress (ESS) and is associated with development and progression of coronary artery disease (CAD). ESS is the parallel frictional force exerted by blood flow on the endothelial luminal surface of the arterial wall. While ESS may not be the only cause of atherosclerosis, it is postulated that low ESS creates a pro-atherogenic environment and high ESS is athero-protective. Computational fluid dynamics (CFD) is an engineering method used to analyse fluid flow and has been increasingly applied to simulate ESS in cardiovascular research. However, current CFD-derived ESS estimates have never been validated and can display significant variability between research groups. Purpose This study aims to provide a comparative analysis of coronary flow using in-silico modelling of coronary flow and particle image velocimetry (PIV). Methods A proximal left anterior descending (pLAD) coronary phantom model was constructed from a patient who had received same day cardiac computed tomography angiography (CTCA) and invasive coronary haemodynamic (Combowire XT) assessment. Patient-specific coronary flow data was applied to CFD boundary conditions and a mock circulatory loop. CFD was validated using PIV under steady flow in a 4-times-scaled model. Same plane velocity field and flow patterns (at mid-luminal of pLAD) from both in-silico and in-vivo data were compared and analysed. Results Mean velocity contours and magnitude were analysed from CFD and PIV. Patient specific average peak velocity obtained from invasive assessment was 0.21m/s. Patient-specific velocity average from CFD was 0.22m/s. The approximated magnitude difference is 4.7%. Both same-axial average cross-sectional velocity estimated by CFD-pLAD and PIV-pLAD are 0.087m/s. Velocity contours and flow patterns simulated in the CFD were also comparable with the PIV results. Conclusion Simulation data from CFD correlate well with experimental PIV results. This early data allows for a non-invasive approach to determine patient-specific coronary haemodynamics (such as velocity profiling and ESS) and form the basis of personalising cardiovascular risk. Insights from this work may aid future development of a novel tool that incorporates CFD-derived ESS estimates with patient-specific risk factors to better predict rapid CAD progression. Funding Acknowledgement Type of funding sources: Public Institution(s). Main funding source(s): Cardiac Society of Australia and New Zealand Monash Institute of Medical Engineering

Validation, verification, and quality control of computational fluid dynamics analysis for indoor environments using a computer‐simulated person with respiratory tract

AbstractComputer‐simulated persons (CSPs) with respiratory systems have been developed for microclimate analysis around the human body and inhalation exposure analysis, for detailed assessment of comfort and health risks in indoor spaces. This study examined and validated the prediction accuracy of a CSP, for precise estimation of indoor environmental quality (IEQ). The flow‐field prediction accuracy was thoroughly examined in a grid analysis using the CSP and a thermal manikin for benchmarking. The model incorporated unsteady breathing and human postural sway, and assessed their impact on the microclimate around the human body. The numerically estimated flow field was validated using experimental particle image velocimetry (PIV) data, with a detailed grid independence test. Considering the practical use of the respiratory tract model for the inhalation exposure risk assessment, the prediction accuracy of particle transport and deposition analysis was examined using previously published in vivo experimental results. This analysis revealed that the impact of transient breathing and body vibrations on the reproduction of the thermal plume around the human body is quite weak; consequently, these conditions can be ignored from the macroscopic perspective of indoor airflow analysis.

Jets and vortices propel fish forward

Time-resolved particle image velocimetry experiments show the generation, evolution, and dissipation of jets and vortices when eel-like fish turn.

The influencing mechanism of iliac vein stent implantation for hemodynamics at the bifurcation

In the intervention with stent implantation for iliac vein compression syndrome (IVCS), it remains unclear about the influencing mechanism of the structure and implantation position of the stent for the hemodynamics of the affected site. In this paper, an iliac vein model was established. Besides, the computational fluid dynamics (CFD) was utilized to analyze the time-averaged wall shear stress (TAWSS) and oscillatory shear index (OSI) in a sine period after stent implantation based on the three different implantation positions of two iliac vein stents (the left branch outlet, contralateral disturbed flow and main iliac vein). The influence of the structure and implantation position of the stent on blood flow was revealed. These findings were verified by the particle image velocimetry (PIV) experiment. The results indicated that the maximum blood flow velocity of the iliac vein decreased after the stent implantation. Among the three positions, the influence of stent implantation on the iliac vein blood flow was the least when the stent implantation was performed at the left branch outlet; the influence of stent implantation on the iliac vein blood flow was the greatest when the stent implantation was performed at the contralateral disturbed flow. Moreover, there was little influence of Venastent implantation on the blood flow. These results could provide a scientific foundation for implantation treatment and stent design related to IVCS.

Evaluation and development of a predictive model for conjugate phase change heat transfer of energy storage system partially filled with porous media

The present study proposes a predictive model to explore the effect of partially filled porous media on the conjugate heat transfer characteristic of phase change material (PCM) with interfacial coupling conditions between pure fluid region and porous region. The enthalpy-porosity method, local thermal non-equilibrium model and Darcy-Forchheimer law are comprehensively considered to describe the convective heat transfer process in porous media. The modified model is then validated by benchmark data provided by particle image velocimetry (PIV) experiments. The phase change behavior, heat transfer efficiency and energy storage performance are numerically investigated for different partial porous filling strategies in terms of filling content, position, height of porous foam and inclination angles of cavity. The results indicate that due to the resistance in porous region, the shear stress exerted by the main vortex (natural convection) in pure fluid region and the momentum transferred, a secondary vortex phenomenon appears in the porous region near the fluid/porous interface. Moreover, such discontinuity of permeability and fluid-to-porous thermal conductivity results in the cusp of phase change interface at the horizontal fluid/porous boundary. Among four partial porous filling cases, the lower porous filling one has more desirable heat transfer performance, and the 3/4H lower porous filling configuration is the best solution for optimization of the latent heat thermal energy storage (LHTES) systems. For tilted cavity, the increase of inclination angle positively affects the heat transfer efficiency as well as the energy storage rate of the LHTES system, where the performance of 3/4H lower porous filling configuration is further highlighted.

Analysis of the NH3 blended ratio on the impinging flame structure in non-premixed CH4/NH3/air combustion

Ammonia as a carbon-free and high-hydrogen fuel with high energy density, is considered a promising alternative fuel. The impinging combustion characteristics of ammonia/hydrocarbon fuel blends are investigated, which can overcome the weaknesses of pure ammonia fuel such as low flammability and unstable combustion. The coupling effect of ammonia blended ratios ranging from 5% to 30% and flame-wall interaction is focused on the non-premixed impinging flame using Particle Image Velocimetry experiments and Large Eddy Simulation. Results show that the flame colors of low NH3 blended ratio are divided into two parts, brighter yellow at the top and blue at the bottom. However, the blue flame disappears and the yellow one gradually takes over the entire flame with the increase of NH3 blended ratios. It is also noted that the attached-wall flame extension distance is shortened with the increase of NH3 blended ratio, which is related with the phenomenon of wall quenching due to the coupling action of the flame-wall interaction and NH3 blended ratios. The phenomenon is attributed to the diffusive and convective instabilities indicated by the vorticity distribution. Besides, the CO2 and CO distributions show that the increase of NH3 blended ratios can effectively reduce the emissions of carbon oxide. Meanwhile, the wide range of high temperature area is formed in 20% NH3 blended ratio case, which promotes the progress of chemical reaction. And then the amount of NO and other key species rises sharply in 20% NH3 blended ratio case in the near wall region, which can consequently affect the flame compositional structure.

Optimized acoustic streaming generated at oblique incident angles to improve ultrasound thrombolysis effect.

Combined with thrombolytic drugs and/or microbubbles, ultrasound (US) has been regarded as a useful tool for thrombolysis treatment by taking its advantages of noninvasive, non-ionization, low cost, and accurate targeting of tissues deep in body. Recently, low-intensity pulsed US, which can cause fewer complications by stable cavitation and acoustic streaming other than more violent effects, has attracted broad attention. However, the thrombolysis effect in practice might not achieve expectation because there is not an ideal parallel multilayered structure between the skin and the targeted vessel. Therefore, the current work aims to better elucidate the influence of US incident angle on the generation of acoustic streaming and thrombolysis effect. Systemic numerical and experimental studies, namely, finite element modeling (FEM), particle image velocimetry (PIV), and in vitro thrombolysis measurements, were performed to estimate the acoustical/streaming field pattern, maximum flow velocity, and shear stress on the surface of thrombus, as well as the lysis rate generated at different conditions. These methods aim at verifying the hypothesis that streaming-induced vortices can further accelerate the dissolution of the thrombus and optimized thrombolysis effected can be achieved by adjusting US incident angles. The pool data results showed that the variation trends of the flow velocity and shear stress obtained from FEM simulation and PIV experiments are qualitatively consistent with each other. There exists an optimal incident angle that can maximize the flow velocity and shear stress on the surface of thrombus, so that superior stirring and mixing effect can be generated. Furthermore, as the flow velocity and shear stress on thrombus surface are both highly correlated with the thrombolysis effect (the correlation coefficient R1=0.988, R2=0.958, respectively), the peak value of lysis rate (increase by at least 5.02%) also occurred at 10°. The current results demonstrated that, with appropriately determined incident angle, higher thrombolysis rate could be achieved without increasing the driving pressure. It may shed the light on future US thrombolysis planning strategy that, if combined with other advanced technologies (e.g., machine-learning-based image analysis and image-guided adaptive US emission modulation), more efficient thrombolytic effect could be realized while minimizing undesired side-effects caused by excessively high pressure.

Analysis Of PIV Images Of Transonic Buffet Flow By Recurrent Deep Learning Based Optical Flow Prediction

In the past few years, several algorithms have been proposed that leverage deep learning techniques within the data analysis workflow of particle-image velocimetry (PIV) experiments. This emerging body of work has shown that deep learning has the potential to match or outperform state-of-the-art classical algorithms in terms of efficiency, accuracy, and spatial resolution. Due to the significant relevance of PIV experiments in the broader fluid mechanics community, progress in PIV processing approaches based on state-of-the-art machine learning tools has a major impact across a range of problems in applied physics and engineering where velocity components of flow fields need to be determined. In contrast to existing methods, these approaches are general, near-automated and yield per-pixel flow estimates. Primary work on deep learning for PIV is promising, but important questions concerning the application to challenging engineering problems remain open and more scientific analyses are necessary to substantiate the general confidence amongst practitioners in such neural network based tools. This motivates us to employ a novel deep learning based PIV approach called RAFT-PIV to a challenging application of a recent research subject. That is, we use our neural optical flow approach to evaluate Background-Oriented Schlieren (BOS) and PIV images obtained by measurements of the transonic flow around a supercritical DRA-2303 profile and demonstrate that it can be used as a direct one-to-one replacement for standard cross-correlation based counterparts.

Coupled Bionic Drag-Reducing Surface Covered by Conical Protrusions and Elastic Layer Inspired from Pufferfish Skin

Inspired by the drag-reducing properties of the cone-like spines and elastic layer covering the pufferfish skin, important efforts are underway to establish rational multiple drag-reducing strategies for the development of new marine engineering materials. In the present work, a new drag-reducing surface (CPES) covered by conical protrusions (sparse "k-type" with rough height k+ = 13-15) and an elastic layer are constructed on copper substrate via a hybrid method, combining the sintering and coating processes. The drag-reducing feature of the prepared CPES biomimetic surface is achieved by rheometer and particle image velocimetry (PIV) experiments. To comprehensively investigate its drag reduction mechanism, the porous copper substrate (PCS), copper substrate (CS), conical protrusion resin substrate (CPRS), and conical protrusion porous copper substrate (CPPCS) were used for a comparative analysis. In laminar flow, we discovered that the conical protrusion structure and wettability of the elastic surface coupling affect the CPES sample's drag-reducing performance (7-8%) and that the interface produced slip to reduce the viscous drag. In turbulent flow, the CPES biomimetic surface exhibits an 11.5-17.5% drag-reducing performance. Such behavior was enabled by two concurrent mechanisms: (i) The conical protrusions as vortex generators enhance the number of vortices and the wake effect, enabling faster movement of downstream strips, reducing viscous drag; (ii) The conical protrusion elements break and lift large-scale vortices to produce numerous small-scale vortices with low energy, effectively weakening perturbations and momentum exchange. Additionally, the elastic layer shows high adhesion and stability on copper substrate after sandpaper abrasion and water-flow erosion tests. The copper substrate surface formed by the sintering method is also covered with dense porous structures, which gives the elastic layer and conical protrusions excellent combined robustness. Our findings not only shed new light on the design of robust drag-reducing surfaces but also provide new avenues for underwater drag reduction in the field of marine applications.

Pressure From Data-Driven-Estimated Velocity Fields Using Snapshot PIV And Fast Probes

The pressure field estimation can be performed from Particle Image Velocimetry (PIV) data on the basis of the Navier-Stokes equation. This requires time-resolved velocity fields to directly compute time derivatives, which are often difficult to achieve due to hardware limitations or cost. Alternatively, snapshot PIV experiments are more affordable but require additional assumptions to model the time derivatives. In this work, we propose the use of data-driven techniques to combine snapshot PIV and high-repetition-rate sensors measurement, and obtain time-resolved estimated velocity fields. The instantaneous pressure fields can thus be computed.We explore building the relation between temporal modes of the flow field and the probe data is built using either linear and non-linear estimators, i.e. Extended Proper Orthogonal Decomposition (Extended POD) and a Multilayer Perceptron (MLP), respectively. The performances of both data-driven methods, as well as Taylor’s hypothesis model for the time derivatives, are tested on the synthetic dataset of a fluidic pinball and on experimental measurements in the wake of a wing. The results show that the data-driven pressure estimations have sufficient accuracy and are not affected by time propagation of the error. Both data-driven methods outperform the model-based method rooted on the Taylor’s hypothesis for the tested cases having compact POD spectra.

Super-Resolution Study Of Wall Sheer Stress In Cerebral Aneurysm With AI PIV

Aneurysms are one of the most common unexplored causes of death. There are many studies conducted by scientists around the world on this issue. The main problem raised is the cause of aneurysms. Here, the division includes several branches, i.e. genetic conditions, lifestyle, eg. smoking, and finally hemodynamic factors of blood flow. It is well known that the aneurysm is formed and ruptured at the site of high wall shear stress (WSS). This paper presents the comparison of results of CFD blood flow simulations with experimental values obtained with the Particle Image Velocimetry (PIV) method in a 3D model of a cerebral aneurysm. A 3D model of a cerebral aneurysm was developed using specific patient CT data. The hemodynamical parameters were analysed for different flow rates in a 3D model under pulsatile flow conditions. The CFD simulation was done using Ansys CFX software. Results of the simulation were validated with the experimental PIV method. Two different PIV methods were employed: traditional PIV based on cross-correlation (CC) algorithms and Artificial Intelligence PIV (AI PIV) based on Deep Learning and Convolutional Neural Networks. The obtained results showed a better correlation between CFD simulation and types of PIV analysis for data of AI PIV, which allows for a much higher spatial resolution of the resulting velocity field. Analysis of the results was promising and showed that AI PIV could be used to get accurate results.

Spatial–temporal evolution of the pressure field generated by a plasma actuator in quiescent air

Dielectric barrier discharge plasma actuators, capable of generating quasi-steady wall jets, are well suited for flow control on various problems. Compared with the detail related to the induced velocity field, there are limited results available for the pressure field created by the plasma actuator. However, the profound merits of understanding the evolution of the pressure field are to reveal the controlling mechanism. Here, the time-averaged and the phase-averaged pressure field distributions are obtained by using a pressure reconstruction method based on the velocity field from particle image velocimetry experiments. According to the discharge regimes, the formation mechanism of the pressure field is discussed. During the streamer discharge stage, the pressure close to the upper electrode is decreased under the influence of the induced heating caused by the high-frequency and high-amplitude pulsed current, leading to the air above the plasma actuator being drawn toward the wall surface. During the glow discharge stage, under the effect of suction generated by the streamer discharge, the pressure near the wall is increased and the plasma actuator generates a favorable pressure gradient, which provides advantageous conditions for the airflow acceleration. During the discharge quenching stage, the effect of the plasma actuator vanishes and the influence of viscous force is strengthened. Therefore, the adverse pressure gradient is gradually formed and the velocity of the wall jet is decreased compared to that of the glow discharge stage. The change of pressure field in a period can be summarized into three processes: pressurization, pressure release, and pressure recovery.

Experimental Study of Collateral Patency following Overlapped Multilayer Flow Modulators Deployment

Decades after its introduction, endovascular aneurysm repair remains a challenging procedure with risks of collateral patency failure. Here, we investigate the ability of a porous stent, the Multilayer Flow Modulator (MFM), to maintain renal perfusion after a single or overlapping case. Silicone models representing an ideal infrarenal AAA geometry were used to analyze and compare three cases (control, single MFM and two overlapped MFMs). Micro-computed tomography was used to image the deployed MFM devices geometry and evaluate pore size and density along with porosity in both two (planimetric) and three dimensions (gravimetric). Laser particle image velocimetry (PIV) experiments were performed to image velocity and vorticity fields at the aorta-renal bifurcation. Flow experiments revealed renal arteries perfusion preservation in both single and overlapped cases. Microstructure analysis revealed an uneven distribution of wires in the MFM devices leading to local change in planimetric porosity and pore size. Overlap of a second MFM device led to a significant decrease in those 2D metrics but did not affect the gravimetric porosity and the branch perfusion. This first microstructure evaluation of MFM device combined with flow experiments revealed the ability of the device to preserve collateral flow thanks to a highly porous microstructure.

Drag reduction in a turbulent boundary layer with sinusoidal riblets

We report on an experimental investigation on the effect of sinusoidal riblets on the near-wall characteristics of a turbulent boundary layer. The investigated riblets are characterized by a fixed wavelength and two different values of the amplitude. We comment on the flow field organization via hot wire anemometry, planar and stereoscopic particle image velocimetry experiments; furthermore, we infer on the friction drag, directly measured with a load cell, comparing the sinusoidal riblets to the reference case of riblets aligned with the mean flow (longitudinal riblets) and the Smooth case.We show that the sinusoidal riblets generally yield higher drag reduction, attaining values as large as 10%, compared with the longitudinal riblets that are limited to 8% under the same conditions. We demonstrate that the drag reduction is associated with an overall attenuation of the turbulence intensity in the buffer layer. Furthermore, we provide statistical evidence of the fact that the sinusoidal riblets are responsible for an attenuation of the Reynolds shear stresses that contribute the most to turbulence production.From the detection of the accelerated events in the buffer layer, we show that the sinusoidal riblets lead to a weakening of the intensity of the events in the streamwise plane and an enhancement of the spanwise induced motion. We relate this mechanism to that responsible for drag reduction when using spanwise wall oscillations, suggesting a possible effect of a secondary alternating vorticity in the grooves of the sinusoidal riblets.

PIV experiments and CFD simulations of liquid–liquid mixing in a planetary centrifugal mixer (PCM)

A new type of blade-free mixer, the planetary centrifugal mixer (PCM), has recently attracted widespread interest. The aim of this study was to simulate the macromixing and micromixing characteristics of two miscible liquids in a PCM. Particle image velocimetry experiments were performed to validate the computational fluid dynamics model. The mixing index and the potential of diffusive mixing based on the intensity and scale of segregation were employed to accurately evaluate the mixing homogeneity. The results showed the remarkable mixing efficiency of the PCM; the mixing process was 95% completed in 10 s in most cases. Chaotic flows and vortices were observed in the PCM, which promoted mixing. After optimizing the design, the mixing time was reduced by 54%, and the extensional efficiency, velocity magnitude, and strain rate were increased by 38%, 35%, and 58%, respectively.

Wake Propagation and Characteristics of a Multi-Rotor Unmanned Vehicle in Forward Flight

In this study, experimental investigations are used to explore the wake propagation and characteristics of a multi-rotor unmanned air vehicle (UAV) in a forward flight mission. Qualitative smoke visualization is used first to gain a qualitative understanding of wake characteristics above and below the body of the multi-rotor UAV which is used as guidance for quantitative particle image velocimetry (PIV) experiments which better resolve the region in the vicinity of the multi-rotor UAV body. The experimental results over a wide range of advance ratios show that as the advance ratio increases, achieved by either lower rotor speeds or higher flight speeds, the distance by which the wake propagates below the UAV is reduced. While above the UAV, the flow returns to the freestream flow closer to the body as the advance ratio increases. Therefore, this study concludes that proximity effects are reduced as the advance ratio increases. Findings from this study can be used to inform in situ sensor placement so that sensor readings are minimally affected by the wake from the multi-rotor UAV. Velocity measurement corrections are provided for sensors mounted above the UAV which can be used to improve sensor data reliability in forward flight. These results can advance autonomous sensing and increase the utility of multi-rotor UAV observations while providing designers and users further guidance to avoid proximity effects.

Toward detecting atherosclerosis using dynamic laser speckle contrast imaging: A numerical study

The disease atherosclerosis causes stenosis inside the patient’s arteries, which often eventually turns lethal. Our goal is to detect a stenosis in a non-invasive manner, preferably in an early stage. To that end, we study whether and how laser speckle contrast imaging (LSCI) can be deployed. We start out by using computational fluid dynamics on a patient-specific stenosed carotid artery to reveal the flow profile in the region surrounding the stenosis, which compares well with particle image velocimetry experiments. We then use our own fully interferometric dynamic light scattering routines to simulate the process of LSCI of the carotid artery. Our approach offers an advantage over the established Monte Carlo techniques because they cannot incorporate dynamics. From the simulated speckle images, we extract a speckle contrast time series at different sites inside the artery, of which we then compute the frequency spectrum. We observe an increase in speckle boiling in sites where the flow profile is more complex, e.g., containing regions of backflow. In the region surrounding the stenosis, the measured speckle contrast is considerably lower due to the higher local velocity, and the frequency signature becomes notably different with prominent higher-order frequency modes that were absent in the other sites. Although future work is still required to make our new approach more quantitative and more applicable in practice, we have provided a first insight into how a stenosis might be detected in vivo using LSCI.

Method for utilizing PIV to investigate high curvature and acceleration boundary layer flows around the compressor blade leading edge

Particle Image Velocimetry (PIV) is a well-developed and contactless technique in experimental fluid mechanics, but the strong velocity gradient and streamline curvature near the wall substantially limits its accuracy improvement. This paper presents a data processing procedure combining conventional PIV and newly developed Mirror Interchange (MI) based Interface-PIV for the measurement of the boundary layer parameter development in the blade leading edge region. The synthetic particle images are used to analyze the measurement errors in the entire procedure. Overall, three types of errors, namely the errors caused by the Window Deformation Iterative Multigrid (WIDIM) algorithm, the discrete data interpolation and integration, and the wall offset uncertainty, comprise the main measurement error. Specifically, the errors due to the discrete data interpolation and integration and the WIDIM algorithm comprise the mean bias, which can be corrected through the error analysis method proposed in the present work. Meanwhile, the errors due to the WIDIM algorithm and the wall offset uncertainty contribute to the measurement uncertainty. Computational fluid dynamics-based synthetic particle flows were generated to verify the newly developed PIV data processing procedure and the corresponding error analysis method. Results showed that the data processing method could improve the accuracy of PIV measurements for boundary layer flows with high curvature and acceleration and even with significant flow separation bubbles. Finally, the data processing method is also applied in a PIV experiment to investigate the boundary layer flows around a compressor blade leading edge, and several credible boundary flow parameters were obtained.

Constraining optimum swimming strategies in plesiosaurs: The effect of amplitude ratio on tandem pitching foils

Identical tandem flippers of plesiosaurs, which are unique among all animals, have been a source of debate regarding the role of hind flippers in their locomotion. Here, inspired by the kinematics of plesiosaur flippers, the effect of the amplitude ratio on the propulsive performance of in-line tandem pitching foils is investigated through a series of particle image velocimetry experiments. Three leader-to-follower amplitude ratios are considered for the foils pitching over a range of 0–2π phase difference. For the first time, it is shown that the amplitude ratio can significantly affect the performance of the hind foil at spacing larger than one chord length. It is found that the thrust generation of the hind foil at the optimum phase difference augments by 130% when it is pitching at the twice angular amplitude of the upstream foil. Although the total performance of the rear-biased and equal amplitude models reaches similar values, thrust production of the hind foil in the equal amplitude model increases only by 23%. By contrast, the performance of the forward-biased model decreases drastically for all phase differences due to the destructive wake–foil interaction of the hind foil. Studying the instantaneous wake–foil interactions, it is found that high thrust generation is associated with the formation of a vortex pair on the suction side of the hind foil, which causes stronger trailing edge vortices to shed with a greater total wake spacing. Finally, through scaling analysis, high-thrust configurations of tandem models are ranked based on the total efficiency of the system.