Computational Fluid Dynamics Calculations Research Articles

Analysis of surimi extrusion behavior during 3D printing by modified Computational Fluid Dynamics (CFD) and quick prediction of printability using machine learning based on texture data

The purpose of this study was to develop a modifying Computational Fluid Dynamics (CFD) model for precise analysis of surimi extrusion behavior and establish a machine learning method for quickly prediction of surimi printability. The 2D rotational axisymmetric model used in this study to replace the 3D/4D physical model, which could improve the simulation efficiency. To improve the accuracy of the CFD simulations, the wall “no-slip” boundary condition assumed in the textbook was replaced with a boundary condition of wall slip. This reduced the error between the simulated results and the measured ones to <1%, and thus calculated required extrusion pressure below 17,034 Pa for printability. The CFD calculation efficiency was improved 33 times than the previously reports by the modifying CFD model, and the simulated results for the required extrusion pressure of inks could be obtained within 10 s. Additionally, a machine-learning method (partial least squares regression model) based on texture data was proposed to quickly predict the required extrusion pressure of surimi to assessed printability. The machine learning model showed a good performance with an R2 > 0.95.

Ship Flow of the Ryuko-maru Calculated by the Reynolds Stress Model Using the Roughness Function at the Full Scale

The k-omega SST turbulence model is extensively employed in Reynolds-averaged Navier–Stokes (RANS)-based Computational Fluid Dynamics (CFD) calculations. However, the accuracy of the estimation of viscous resistance and companion flow distribution for full-sized vessels is not sufficient. This study conducted a computational analysis of the flow around the Ryuko-maru at model-scale and full-scale Reynolds numbers utilizing the Reynolds stress turbulence model (RSM). The obtained Reynolds stress distribution from the model-scale computation was compared against experimental measurements to assess the capability of the RSM. Furthermore, full-scale computations were performed, incorporating the influence of hull surface roughness, with the resulting wake distributions juxtaposed with the actual ship measurements. The full-scale calculation employed the sand-grain roughness function, and an optimal roughness length scale was determined by aligning the computed wake distribution with Ryuko-maru’s measured data. The results of this study will allow for the direct performance estimation of full-scale ships and contribute to the design technology of performance.

Aerodynamic interference analysis and modeling of heterogeneous rotors

During closed-range rotorcraft missions, strong aerodynamic disturbances occur between multiple aircraft, thus affecting the stability and safety of cooperative flight processes. However, owing to the numerous control variables in rotor models, the cost and complexity of experiments and body-fitted grid simulation methods are high, thereby resulting in low analysis efficiency and difficulty in establishing a highly applicable aerodynamic-interference model. In this study, the effectiveness of a two-dimensional quasi-steady momentum source method for calculating the aerodynamic interference of heterogeneous rotors is validated using the sliding-mesh method. Using Latin sampling and momentum source methods to obtain aerodynamic-interference data points, a multiparameter-coupled heterogeneous-rotor aerodynamic-interference model is proposed and validated. The results indicate that the proposed aerodynamic-interference model provides clear physical significance and demonstrates high accuracy under different rotor configurations. In validation samples, the correlation coefficient between the predicted aerodynamic loss on the lower rotor and the computational fluid dynamics (CFD) calculation results is approximately 0.99, with the maximum errors in predicting thrust and moment losses are less than 7% and 2%. Meanwhile, the correlation coefficient between the predicted values on the upper rotor and the CFD calculation results is 0.82, with the maximum errors in predicting the dimensionless thrust and moment losses are less than 5.5% and 1%. The rapid aerodynamic-interference analysis model proposed herein can serve as a reference for the safe-flight strategy of rotorcraft during close-range cooperative missions.

Aerodynamic Modeling of a Missile Combining Transfer Learning and Dendritic Net

During the initial stages of missile design, finding an efficient method for analyzing aerodynamic characteristics is crucial. This paper proposes a novel aerodynamic modeling method based on a limited computational fluid dynamics (CFD) dataset, combining transfer learning (TL) with Dendritic Net (DD). Initially, we employ CFD-calculated aerodynamic data to establish a pretrained model using DD. Subsequently, the model is adapted to the target domain by TL, predicting aerodynamic parameters under specific conditions. The overall aerodynamic parameters are utilized to generate a relational spectrum through DD’s white-box features, from which primary features are extracted to establish an aerodynamic polynomial model. Finally, the model’s practicality is validated by ballistic flight simulations. The innovation lies in leveraging DD to generate a relational spectrum of aerodynamic parameters, leading to a high-precision polynomial model. Research shows DD outperforms traditional cell-based networks in predicting aerodynamic parameters, and TL reduces the CFD computation workload in the target domain by 3/4 while maintaining prediction accuracy. The polynomial model exhibits superior accuracy compared to the empirical fitting formulas. The method reduces the computational workload for aerodynamic data collection, and through system identification, a high-precision polynomial model is obtained, which provides a reliable basis for missile controller design.

Impact of Blade Modifications on the Performance of a Darrieus Wind Turbine

Vertical axis wind turbines (VAWTs) are gaining increasing significance in the realm of renewable energy. One notable advantage they possess is their ability to operate efficiently in diverse wind conditions, including low-speed and turbulent winds, which are often prevalent in urban areas. In this study, dimples and pitch angles into the rotor blades are used to enhance the aerodynamic performance of a straight-bladed Darrieus turbine. To simulate the turbine’s rotation under transient conditions, computational fluid dynamics calculations are conducted in a two-dimensional setting. The unsteady Navier–Stokes equations are solved, and the k-ω SST turbulence model is employed to represent turbulent flow. The results of the simulation demonstrate that the application of a circular dimple on the pressure side of the blades, positioned at 0.25 of the chord length with a diameter of 0.08 chord length, leads to a 5.18% increase in the power coefficient at λ = 2.7, in comparison to a turbine with plain airfoils. Moreover, when an airfoil with both a dimple and a + 1° pitch angle is utilized, the turbine’s performance at λ = 2.7 improved by 7.17% compared to a plain airfoil, and by 1.8% compared to a dimpled airfoil without a pitch angle. Additionally, the impact of a double dimple on both the pressure and suction sides of the airfoil on turbine performance was investigated. It was discovered that the double-dimpled airfoil exhibited lower performance in comparison to a plain airfoil. The study showed that the utilization of both dimples and pitch angles for airfoils of a Darrieus turbine blade increases the power generated by the turbine.

Research on vibration characteristics of cone valve based on steady state hydrodynamics

The steady-state fluid dynamic force of the cone valve core reduces the stability of the cone valve, thereby affecting its vibration characteristics. This study first derived the formula for calculating the steady-state fluid dynamic force of the valve core using the momentum theorem. Then, the study used the Computational Fluid Dynamics (CFD) method to explore the variation of steady-state fluid dynamic force with the opening degree of the cone valve and the pressure difference at the inlet and outlet. The study also analyzed the steady-state fluid dynamic force of the improved structure cone valve. Combining the CFD calculation results with the established axial vibration and dynamic models of the valve core spring system, a system dynamics simulation model was developed. Innovatively, the study combined experimental and simulation methods to investigate the vibration characteristics of the cone valve and the impact of steady-state fluid dynamic force on these characteristics. The experimental results aligned well with the simulation results, showing that the influence of steady-state fluid dynamic force on vibration characteristics is similar to that of spring force. At a small opening degree, the compensation structure of steady-state fluid dynamic force can reduce axial vibration amplitude. Increasing the inlet pressure from 2.0 MPa to 2.5 MPa results in a decrease in axial amplitude and an increase in the numerical value of the vibration balance position. Increasing the outlet pressure from 0.5 MPa to 0.9 MPa initially decreases axial amplitude and axial waveform factor, followed by an increase. Increasing the spring preload from 9.4 mm to 13.4 mm leads to a decrease in the numerical value of the balance position, a reduction in amplitude, and a decrease in the waveform factor of the axial vibration of the valve core. The study reveals systematic effects of inlet pressure, outlet pressure, and spring preload on the vibration characteristics of the cone valve. The research results in this paper can provide theoretical support for the structural design of hydraulic valves, reduce the impact of fluid dynamics on vibration, and thereby improve the stability of hydraulic system operation.

Experimental and numerical investigations on micromixing performance of multi-orifice cross-flow jet mixers

Multi-orifice cross-flow jet mixers (MOCJMs) are used in various industrial applications due to their excellent mixing efficiency, but few studies have focused on the micromixing performance of MOCJMs. Herein, the flow characteristics and micromixing performance inside the MOCJM were investigated using experiments and computational fluid dynamics (CFD) simulations based on the Villermaux/Dushman system and the finite rate/modified eddy dissipation model (FR/MEDM). The optimal A value was correlated with the characteristic parameters of MOCJMs to develop a CFD calculation method applicable to the study of the micromixing performance of the MOCJMs. Then the micromixing efficiency was evaluated using the segregation index XS, and the effects of operational and geometric parameters such as mixing flow Reynolds number (ReM), flow ratio (RF), total jet area (ST), number of jet orifices (n) and outlet configuration on the micromixing efficiency were investigated. It was found that the intensive turbulent region generated by interactions between jets, as well as between jets and crossflows, facilitated rapid reactions. XS decreased with increasing ReM and decreasing RF. Furthermore, MOCJMs with lower ST, four jet orifices, and the narrower outlet configuration demonstrated a better micromixing efficiency. This study contributes to a deeper understanding of the micromixing performance of MOCJMs and provides valuable guidance for their design, optimization, and industrial application.

Flow fields prediction for data-driven model of 5 × 5 fuel rod bundles based on POD-RBFNN surrogate model

The fuel rod bundles are a crucial component of nuclear reactors. Research the flow characteristics of the fuel rod bundles can be computationally expensive, despite some progress made in the past few decades. In this paper, the fast prediction of flow fields for 5 × 5 fuel rod bundles is implemented based on a data-driven algorithm. First, a refined model of the 5 × 5 fuel rod bundles with spacer grid is established, and meshes are generated using a hybrid meshes technology. Then, the flow characteristics of the 5 × 5 fuel rod bundles are extensively analyzed using Computational Fluid Dynamics (CFD) methods. A comparison between the CFD results and experimental data reveals a strong agreement, thereby validating our research findings. Subsequently, a reduced-order model (ROM) called the proper orthogonal decomposition (POD)-radial basis function neural network (RBFNN) surrogate model is proposed. the POD algorithm is utilized to extracting the dominant flow modes, and the RBFNN is employed to predict the POD mode coefficients of non-sample points. By linearly combining the predicted POD mode coefficients with the POD modes, rapid prediction of flow fields for non-sample points can be achieved. Compared to the full-scale CFD simulations, it finds that the proposed POD-RBFNN surrogate model not only significantly improved efficiency, but also maintained a high level of predictive accuracy. It is believed that the research results hold important value for CFD calculations within nuclear reactor cores.

Rapid data-driven individualised shape design of AUVs based on CFD and machine learning

ABSTRACT The shape design of autonomous underwater vehicles (AUVs) has an essential influence on their hydrodynamic performance. In this study, a data-driven approach is proposed based on computational fluid dynamics (CFD) and machine learning to efficiently design the hydrodynamic shape of AUVs with different sailing velocities, angles of attack (AOAs) and volumes. Based on the data-driven approach, a shape optimisation design for AUVs is established. Moreover, the variation of optimal shape with the design constraints of AOA, velocity and volume is also explored, and the results prove that a relatively larger body with a sharper head is more beneficial for the Myring shape AUVs sailing with a higher velocity and a larger AOA. The results of CFD calculations and those of the data-driven approach are compared, and a circulating flume test is carried out to prove the reliability of CFD method. The comparison verifies the reliability of the optimisation result.

A residual graph convolutional network for setting initial flow field in computational fluid dynamics simulations

The computational cost of computational fluid dynamics (CFD) simulation is relatively high due to its computational complexity. To reduce the computing time required by CFD, researchers have proposed various methods, including efficient time advancement methods, correction methods for discrete control equations, multigrid methods, reasonable initial field setting methods, and parallel methods. Among these methods, the initial field setting method can provide significant performance improvements, but there is little work on it. Existing CFD industrial software typically uses inflow conditions for the initial flow field or applies empirical methods, which can cause instability in the CFD calculation process and make convergence difficult. With the rapid development of deep learning, researchers are increasingly attempting to replace CFD simulations with deep neural networks and have achieved significant performance improvements. However, these methods still face some challenges. First, they can only predict the computational flow field on regular grids. They cannot directly make predictions for irregular grids such as multi-block grids and unstructured grids, so the final flow field can only be obtained through interpolation and similar methods. Second, although these methods have been claimed to provide high accuracy, there is still a significant gap in performance with CFD and they cannot yet be applied to real scenarios. To address these issues, we propose a Residual Graph Convolutional Network for Initial Flow Field Setting (RGCN-IFS) in CFD simulations. This method converts the grid into a graph structure and uses an improved graph neural network to predict the flow field. In this way, we can predict the flow field on any type of grid. More importantly, this method does not directly replace CFD simulations, but it rather serves in an auxiliary role, providing appropriate initial flow fields for the CFD calculations, improving the convergence efficiency while ensuring calculation accuracy, and directly bridging the accuracy gap between intelligent surrogate models and CFD simulations.

Equilibrium contact angles and dewetting in capillaries

In this work, we extend the model of contact angles that we have previously developed for sessile drops on a wetted surface to the case of a meniscus in a capillary. The underlying physics of our model describe the intermolecular forces between the fluid and the surface of the capillary that result in the formation of a thin, non-removable fluid layer that coats the capillary wall. We describe the shape of the meniscus using a Young–Laplace equation and an incompressible, two-phase, computational fluid dynamics (CFD) calculation, both modified to take into account intermolecular forces using the disjoining pressure model. We find that our numerical solutions of the Young–Laplace equation and equilibrium meniscus shapes obtained by CFD agree well with each other. Furthermore, for capillaries that are sufficiently larger than the thickness of the non-removable film, our numerical solutions agree well with the effective contact angle model that we previously developed for sessile drops. Finally, we observe that it is possible to tune the disjoining pressure model parameters so that the intermolecular forces between the liquid and solid molecules become so strong compared to the surface tension that our formula for effective contact angle gives an imaginary solution. We analyze this situation using CFD and find that it corresponds to dewetting, where the bulk liquid detaches from the walls of the capillary leaving behind the non-removable thin liquid film.

Computational Fluid Dynamic study of gas mixtures in a Non-Thermal Plasma reactor for CO2 conversion with Argon as diluent gas

CO2 utilization has been an emerging technology of increasing global interest due to its direct impact in limiting greenhouse gas emissions. In this contribution, the fluid dynamic behavior of a CO2 conversion non-thermal plasma (NTP) in a dielectric barrier discharge (DBD) reactor is studied through computational fluid dynamics (CFD) simulations.&#x0D; Calculations are provided in conjunction with experimental results and the thermodynamic characterization of the compounds and mixtures involved. This CFD study utilizes a well-established methodology that allows the optimization of fluid flow with limited computational burden.&#x0D; Firstly, results are presented for an Example Case, in which several variables are studied both at the final iteration as well as across iterations. Secondly, a range of Study Cases, changing the inlet composition and volume rate, are presented. Average velocity is one of the most significant variables to predict the reactor’s yield, while the temperature, density and pressure in the reactor remain, in most cases, almost constant.&#x0D; The resulting CFD computations describe the behavior of the fluids in the reactor in a predictive manner for future experimental results. Limitations in the fluid’s characterization occur due to not explicitly including the plasma reaction, which will be aimed at in future contributions.

Nominal wake field of a ship with moonpool

ABSTRACT To investigate the influence of a moonpool on the nominal wake field of a ship during calm water navigation, computational fluid dynamics calculations are performed using the STAR-CCM + software. The analysis found that the influence of boundary layer separation in the moonpool causes the hull boundary layer thicker, which reduces the average axial velocity of the propeller behind the moonpool and enhances the wake. The uniformity of the axial velocity distribution at the propeller disc is also reduced, worsening the uniformity of the wake field. In addition, the wake field characteristics of rear propellers change periodically with the fluctuation of moonpool. As a result the propeller bearing force and vibration characteristics change periodically with not only blade frequency and multiple blade frequency, but also the fluctuation of moonpool. These phenomena make the performance of propellers on ships with moonpool different from that of normal ships.

EMPIRICAL MODELS’ APPLICABILITY FOR CALCULATING THE CENTRIFUGAL PUMP'S IMPELLERS GEOMETRIC PARAMETERS

This paper deals with the comparison of empirical computational models used to achieve the geometrical parameters of pumps impellers with models based on Computational Fluid Dynamics (CFD) simulations and optimization algorithms. Also, the actuality of the empirical models application in modern pump manufacturing industry is analyzed. Further, an empirical model for calculating the pump impeller geometrical parameters was presented. Applying this model, the calculation of the centrifugal pump impeller parameters for CH 6,3/20 1,1-2 canned motor pump was performed. The procedure for parameterization and generation of the geometric model based on parameters obtained from this model through ANSYS DesignModeler was also presented. CFD simulation based on ANSYS CFX was used to obtain the operating characteristic of the obtained centrifugal pump impeller. The authors carried out the comparison of the results of the designed model with the original one and with the optimized impeller obtained using optimization algorithms and CFD calculations.

Development of a Novel Water Jet Polisher Using Soft Abrasives for Small Complex-Structure Heat Pipes of Aluminum Alloy Produced Using Additive Manufacturing.

It is a challenge to polish the interior surface of a small bent pipe with complex structures and sizes less than 0.5 mm. This is because of the fact that traditional polishing methods could destroy, block, or break the small complex structures. For a small bent pipe made of aluminum alloy produced using additive manufacturing, the defects, such as adhered powders and spatters, are easy to jam the pipe without polishing, possibly resulting in catastrophic failure for aerospace applications. To overcome this challenge, a novel water jet polisher was developed using soft polymethyl methacrylate (PMMA) abrasives. After polishing a specific area, the adhered powders on the interior surface were reduced from over 140 to 2, 3, and 6 by the soft abrasives with mesh sizes of 200, 400, and 600, respectively. The surface roughness Sa was decreased from 3.41 to 0.92 μm after polishing using PMMA abrasives with a mesh size of 200. In comparison, silica abrasives were also employed to polish the small bent pipes, leading to the bent part of pipes breaking. However, this kind of failure was absent when using soft abrasives. Computational fluid dynamics calculations elucidate that a peak erosion rate of silica abrasives for a bent pipe with a turn angle of 30° is 2.18 kg/(m2·s), which is 17 times that of soft abrasives. This is why the small bent pipe was broken using silica abrasives, whereas it remained intact when polished with soft abrasives. In addition, water jet polishing has a lower erosion rate, a relatively smooth erosion curve, and less erosion energy, leaving the bent parts intact. The developed soft abrasive water jet polisher and the findings of this study suggest new possibilities for cleaning the adhered powders and spatters and polishing the interior surface of small bent pipes with complex structures.

Numerical study of novel OME1−6 combustion mechanism and spray combustion at changed ambient environments

AbstractFor a climate-neutral future mobility, the so-called e-fuels can play an essential part. Especially, oxygenated e-fuels containing oxygen in their chemical formula have the additional potential to burn with significantly lower soot levels. In particular, polyoxymethylene dimethyl ethers or oxymethylene ethers (PODEs or OMEs) do not contain carbon-carbon bonds, prohibiting the production of soot precursors like acetylene (C2H2). These properties make OMEs a highly interesting candidate for future climate-neutral compression-ignition engines. However, to fully leverage their potential, the auto-ignition process, flame propagation, and mixing regimes of the combustion need to be understood. To achieve this, efficient oxidation mechanisms suitable for computational fluid dynamics (CFD) calculations must be developed and validated. The present work aims to highlight the improvements made by developing an adapted oxidation mechanism for OME1−6 and introducing it into a validated spray combustion CFD model for OMEs. The simulations were conducted for single- and multi-injection patterns, changing ambient temperatures, and oxygen contents. The results were validated against high-pressure and high-temperature constant-pressure chamber experiments. OH*-chemiluminescence measurements accomplished the characterization of the auto-ignition process. Both experiments and simulations were conducted for two different injectors. Significant improvements concerning the prediction of the ignition delay time were accomplished while also retaining an excellent agreement for the flame lift-off length. The spatial zones of high-temperature reaction activity were also affected by the adaption of the reaction kinetics. They showed a greater tendency to form OH* radicals within the center of the spray in accordance with the experiments.

Flow fields prediction for data-driven model of parallel twin cylinders based on POD-RBFNN and POD-BPNN surrogate models

Flow around cylinders is an important phenomenon in many different engineering fields. In this paper, the fast prediction of the pressure fields of parallel twin cylinders is implemented based on a data-driven algorithm. Firstly, the pressure fields of parallel twin cylinders with a low Reynolds number are obtained through the Computational Fluid Dynamics (CFD) method. The pressure fields at different time steps are collected to form a snapshot matrix. The Proper Orthogonal Decomposition (POD) algorithm is then applied to obtain the POD basis vectors of the snapshot matrix, enabling the reconstruing the pressure fields. Subsequently, two reduced-order models (ROM) called the POD-RBFNN and POD-BPNN surrogate models are proposed in this paper. The POD-RBFNN surrogate model uses the Radial Basis Function Neural Network (RBFNN) to train the POD mode coefficients obtained from the POD algorithm, while the POD-BPNN surrogate model uses the Backpropagation Neural Network (BPNN) for the same purpose. Linearly combining the POD mode coefficients predicted by the POD-RBFNN or POD-BPNN surrogate models with the POD basis vectors obtained from the POD algorithm enables fast and efficient prediction of pressure fields for non-sample points. Finally, comparisons are made between the predicted pressure fields obtained from these two surrogate models and the actual values obtained through CFD simulations. It is found that both the POD-RBFNN and POD-BPNN surrogate models proposed in this paper not only significantly improve efficiency but also maintain a high level of accuracy. However, the training time of the POD-RBFNN surrogate model is significantly shorter than that of the POD-BPNN surrogate model. Additionally, the POD-RBFNN surrogate model exhibits smaller Root Mean Square Errors (RMSE) and Mean Absolute Error (MAE). For the data-driven model of parallel twin cylinders described in this paper, the POD-RBFNN surrogate model is more suitable to predict the pressure fields. The research results in this paper are believed to hold significant value for CFD calculations of parallel twin cylinder models, offering essential guidance for a deeper understanding of cylinder flow problems and optimizing engineering design.

Development of advanced hybrid computational model for description of molecular separation in liquid phase via polymeric membranes

Separation of molecules using membranes has been computationally studied in this work for understanding the separation process and also assessment of integrating machine learning (ML) techniques to computational fluid dynamics (CFD) techniques. We initially performed CFD computations to simulate mass transfer in a membrane-based molecular separation system, and then exported the mass transfer data for training ML models. Indeed, a hybrid model was developed and employed for membrane process in separating a solute from a solution. We have modeled a data set containing more than 14 thousand data rows, which has two inputs r (m) and z (m) and single response, i.e., C (mol/m3) that is solute’ content in the solution. Decision tree-based models are used in this modeling. Simple Decision Tree (DT), Boosted Decision Tree (BDT) that is DT alongside AdaBoost, and Random Forest (RF) are the selected models for this research. The ML techniques were further optimized via a novel HHO algorithm. Final assessments show the R2 scores of 0.9991, 0.9874, and 0.9997 respectively for BDT, DT, and RF models. Based on this result and other analysis, RF is selected as the best method in this study for simulation of membrane process. This model has an RMSE error rate of 2.22 × 103.

Tackling Modern Sailing Challenges with a CFD-based Dynamic VPP

A dynamic Velocity Prediction Program (VPP) integrated in a Computational Fluid Dynamics (CFD) code is described. Aerodynamic forces are obtained either through empirical coefficients or interpolated from aerodynamics matrices. These aerodynamic forces are then input to the hydrodynamics CFD solver, which solves both the flow and the motions of the boat, resulting in a closely coupled VPP. For a given True Wind Angle and True Wind Speed a sail power parameter is optimised to obtain the best possible boat speed within heel angle constraints. This approach allows naval architects to swiftly and precisely compare several yacht designs in real sailing configurations using only a few CFD computations. Several advanced features recently added to this program are covered in this paper including convergence criteria, automatic grid refinement, foil fluid-structure interaction, multiple aerodynamics models and rudder control. Results obtained from our CFD VPP on a 40-feet fast-cruising yacht demonstrates promising agreement with other existing VPP polars, affirming the accuracy and reliability of our approach. The CFD VPP presented was also successfully applied to an IMOCA, a 60-feet racing yacht.

Advancing fluid dynamics simulations: A comprehensive approach to optimizing physics-informed neural networks

Flow modeling based on physics-informed neural networks (PINNs) is emerging as a potential artificial intelligence (AI) technique for solving fluid dynamics problems. However, conventional PINNs encounter inherent limitations when simulating incompressible fluids, such as difficulties in selecting the sampling points, balancing the loss items, and optimizing the hyperparameters. These limitations often lead to non-convergence of PINNs. To overcome these issues, an improved and generic PINN for fluid dynamic analysis is proposed. This approach incorporates three key improvements: residual-based adaptive sampling, which automatically samples points in areas with larger residuals; adaptive loss weights, which balance the loss terms effectively; and utilization of the differential evolution optimization algorithm. Then, three case studies at low Reynolds number, Kovasznay flow, vortex shedding past a cylinder, and Beltrami flow are employed to validate the improved PINNs. The contribution of each improvement to the final simulation results is investigated and quantified. The simulation results demonstrate good agreement with both analytical solutions and benchmarked computational fluid dynamics (CFD) calculation results, showcasing the efficiency and validity of the improved PINNs. These PINNs have the potential to reduce the reliance on CFD simulations for solving fluid dynamics problems.