| Particulate-fluid interaction systems are ubiquitous in various industrial pro-cesses whereas present complex momentum or heat transfer characteristics.Most of the particles in these processes are non-spherical particles(such as,ellipsoid-shaped,cylinder-shaped,tablet-shaped,etc),non-spherical particulate-fluid inter-action systems have its own characteristics and difficulties,which differ markedly from spherical part icles.The unique geometry of non-spherical particles introduces more uncertainties than spherical particles in particle-particle or particle-fluid in-teractions for flow structure and mechanism,therefore,the multi-scale structure is more complex than the case of spherical particles in the multi-phase flow and heattransfer process,and there are still many problems worthy of further study.In this work,firstly,a 3D IB-LBM parallel numerical simulation platform for hot sta-tionary particle-fluid coupling system is established,which is based on CPU-GPU heterogeneous computers.Secondly,the LBM-IBM-DEM numerical simulation platform for dynamic evolution process of dense fluid-particle system is established,and then the magnetic particles-fluid phase coupling in two-phase fluid modelling is carried out.The main research findings and innovations are as follows.A CUDA based 3D IB-LBM is constructed for the multi-scale modeling of forced convection for a class of stationary ellipsoid particle immersed in cool fluid,and the characters of force evolution and heat transfer are revealed.Firstly,the stability and convergence of the algorithm are verified by simulating the problem of a single hot spherical particle under different grids resolution.Furthermore,the numerical studies of a sphere/ellipsoid under different Reynolds numbers are presented,the results are consistent with the literatures,and the correctness of the 3D parallel platform is verified.Secondly,typical non-spherical particle with different sizes and incident angles with different Reynolds numbers were tested,the effects of operating parameters on the drag coefficient and averaged Nusselt number of a single hot stationary non-spherical particle were obtained.Based on the numerical dates,new correlations considering the effect of shape,orientation,and the Reynolds number were developed for predicting the drag coefficient and averaged Nusselt number of non-spherical particles.The correlation are in ranges of the Reynolds number 10 ? Re ? 200,aspect ratio 0.25 ? Ar ? 2.5 and incident angle 0° ? ?? 90° presented,with averaged deviations between the numerical simulation results and the calculated values of the correlation are ?Cd=2.1%and?Nu=1.4%,which are convenient to be used in the macro scale modelling such as in the Eulerian-Lagrangian coupling framework.Based on the parallel numerical simulation platform mentioned above,the particle shape is extended to the asymmetric ellipsoidal particle,and its application is expanded.Through numerical simulation of ellipsoidal particles with different Reynolds numbers and incident angles,the extended simulation platform is further verified.Then,the evolution of drag coefficient and averaged Nusselt number are investigated numerically with the Reynolds number 25 ? Re ? 200,aspect ratio 2 ? Arl,Ar2 ? 6,incident angle 0° ? ? ? 90°,and the drag mechanism and heat transfer characteristics of asymmetrical ellipsoidal particles with the variation of operating parameters are revealed.The numerical simulations are compared in detail with the value calculated from the correlation in the literatures.Based on those dates,two kinds of new correlations are presented,which can be directly used in the calculation of Cd and Nu with the average errors are 3.875%and 1.598%respectively,especially the second one.Based on the parallel numerical simulation platform of single particle,the number of particle is extended from one to multi-particles system,which further expands the scope of application.Then the drag evolution and heat transfer char-acteristics of two hot non-spherical particles in tandem arrangement are studied in detail.In addition,the influence of the inter particle distance and relative in-cidence angles between the two tandem spheroids as well as the multi-particles system on the momentum and heat transfer are studied.A LBM-IBM-DEM numerical simulation platform for dynamic evolution pro-cess of dense particles system is established,numerical simulations of fluid-particle interactions with uniform magnetic fields are implementation effectively.Firstly,the simulation of dry dam break test,the Drafting-Kissing-Tumbling problem of non-magnetic particles and the interaction between magnetic particles are numer-ically simulated,which verifies the correctness of the simulation platform.Based on it,the settling of two magnetic particles in vertical square cavity under exter-nal magnetic field is studied in detail,the settling velocity and the trajectory of particle motion under different working conditions are obtained,and the effects of distance between particles and intersection angle between connection line and direction of magnetic field for the motion behavior of particles are revealed.Next,the effects of different applied magnetic fields on the dynamic behavior of parti-cle groups in fluid particle dense systems are numerically studied,the model can capture the motion trajectory of magnetic particles and the dynamic evolution of chain structure outstandingly when compared with the experimental or numerical simulation results available in the literature.And then,the effect of the volume fraction of magnetic particles on the motion of the particle group is also studied,the multi-scale motion mechanism of magnetic particles in the fluid is revealed from the view of numerical simulations. |
PDF Full Text Request