Welcome to the Multiphase Flow Simulation Laboratory

      Multiphase flow is a flow of several phases. Examples include two-phase flows of gas-solid, gas-liquid or liquid-solid, and three-phase flows of gas-liquid-solid. It has direct applications in many industrial processes including riser reactors, bubble column reactors, fluidized bed reactors, dryers, and particle separators. Many environmental flows and biological flows are also multiphase flows. Our lab is primary dedicated to developing numerical simulation models for multiphase flows. An overriding objective of our research is to improve our understanding of many multiphase flows utilizing computer simulations. We also conduct experimental research to validate our numerical models. Our research is mainly supported by the National Science Foundation (NSF) and Department of Energy (DOE).

The main research topics in our lab are:

  • Direct numerical simulation (DNS) of heat, mass and momentum transfer between particles and fluids
  • Modeling of particle segregation, clustering and agglomeration in particulate flows
  • Developing constitutive equations/closures for two-fluid model such as MFIX
  • Investigating the effect of particle shape on its interaction with surrounding fluid


  Research in Our Lab

 Developing Constitutive Equations for Two-Fluid Models

Particulate flows have a wide range of applications, such as sedimentation process, fluidized bed reactors, and pneumatic conveying. There are two different approaches in modeling particulate flows: Eulerian-Lagrangian approach and Eulerian-Eulerian approach or two-fluid model (TFM). In the Eulerian-Lagrangian approach the particles are tracked individually and their dynamics are solved by the Lagrangian method. In the TFM, the large number of particles (solid phase) are considered to constitute a continuum phase or a virtual fluid; the motion of the solid phase is described by the momentum transport equations similar to the Navier-Stokes equations. Most engineering applications require the use of TFM for better understanding and predicting the performance because of their large scales in terms of the number of particles and the size of devices. However, the primary challenge of two-fluid modeling is the development of constitutive equations for solid phase. For example, it is well known that most real fluids, like air or water, can be considered a Newtonian fluid with a viscosity that is well defined and measurable. However, the determination of viscosity of solid phase remains an open problem; furthermore, the no-slip boundary condition commonly used for viscous fluid may not be applicable to the solid phase.Our approach is to use more detailed Eulerian-Lagrangian approach (e.g., direct numerical simulation (DNS) or discrete-element method (DEM)) and explore its rich simulation data with statistical analysis to develop constitutive equations and boundary conditions for less detailed TFM.

Side-by-side snapshot comparison of DNS and TFM simulations of particle sedimentation. Fluid is colored by white in DNS and by red in TFM. Particle and fluid density ratio is 1.1; initial porosity is 0.38. In this case, DNS is used to test the solid phase viscosity model and boundary condition scheme.

 Low-Cost Method of Treating Flowback Water from Hydraulic Fracturing

In a combined effort, we are collaborating with researchers at SwRI to investigate a cost-effective alternative to remove contaminants found in industrial wastewater utilizing biochar. The biomasses used in the investigation are oat hull, pine wood, oak wood which are used to produce biochar. Biochar is produced by placing the biomass under a process known as pyrolysis. Pyrolysis is the process of uniformly heating biomass in an oxygen deprived environment. The biomass produce in this study is placed in a nitrogen environment and heated in a furnace at multiple temperatures for three hours once steady state is reach.
The main objective of this research is to identify the most effective way and process to remove contaminants found in flowback water, and to investigate how effective each char is for a particular contaminant and how much amount of contaminants can be removed for a given amount of char. Another objective is to determinate model parameters (e.g., adsorbing rate, diffusion coefficient, etc al.) to be used for modeling and predicting the contaminant-removal process in a large scale application.

(a) Biochar production (b) Stagnant biochar and contaminated solution (diffusion process) (c) Mechanical mixing biochar and contaminated solutions (convection and diffusion process ) (d) Flow contaminated solution through biochar packed bed column (convection dominated process).

 Multiscale Modeling of Nanoparticle Clustering

Solvents doped with relatively low concentrations of nanoparticles (with diameters up to 100 nm) are termed as nanofluids. Nanofluids are considered to be attractive as heat transfer fluids and thermal energy storage for concentrated solar power applications. However, nanoparticles tend to agglomerate or clustering, resulting large size aggregates that cause nanofluids unstable (settling) and lose nanoparticle unique properties. How to achieve the desired stability of nanofluids is one of the main challenges in its thermo-fluid related applications.We are currently developing a multiscale approach to study the drivers that cause or prevent nanoparticle clustering. We are investigating the significance in the contribution of a range of forces across different time and length scales. The simulation results provide an estimate for the time scale for the agglomeration and the resultant structure of the agglomerated ensemble of nanoparticles. Subsequently simulations are performed using this numerical model corresponding to the available experimental data in the literature. The predictions from the numerical simulations show that the change in zeta potential (determined in part by the pH of the solvent phase) is a crucial parameter that affects the level of agglomeration of the nanoparticles.
Figure above shows the gradual formation of large agglomerated structures within the medium. The nanoparticles are polystyrene spheres with a diameter of 50 nm and a zeta potential of -100 mV. Water at standard temperature and pressure is considered as the solvent phase. The computational region is set at 1.35 um on both sides, and randomly populated with 500 nanoparticles. A small time step for DLVO force is set at 10E-11 s, and a large time step for fluid force is set at 10E-10 s. A periodic boundary condition is enforced on all four sides. The resultant DLVO force is the primary factor for agglomeration. When the attractive Van der Waals force exceeds the repulsive double layer force, the particles will tend to agglomerate.

 Biomass Pyrolysis in a Fluidized Bed Reactor

Fast pyrolysis is one of the primary technologies to harvest energy from biomass. To facilitate biomass pyrolysis, a fluidized bed reactor is used because of its high mixing efficiency and ease of maintenance. In a fluidized bed reactor, silica sand is preheated to a high temperature (300C-600C), then nitrogen is injected through a distributor at the bottom of the bed to fluidize or lift the sand particles; at the same time, biomass particles are fed into the reactor. Violent collisions between biomass particles and sand in the high temperature and oxygenless environment cause biomass to be rapidly converted to vapors, gases and charcoal. After cooling and condensation, some of these vapors and gases turn into dark-brown bio-oil. However, particles in a biomass fluidized bed have very unique properties and characteristics due to their large density and size differences. Issues such as particle segregation and agglomeration are commonly encountered; they deteriorate the efficiency of biomass pyrolysis in the reactor.

To achieve the highest yield of bio-oil production, it is important to understand the hydrodynamics of particle-fluid flows in the bed reactor and find the optimal operating conditions. The goal of our research is to identify the optimal operating conditions and suitable configurations of biomass fluidized bed reactors for producing the highest yield of bio-oil.

fluidization of particles in a reactor. high resolution movie

click here for a higher resolution video clip
We are currently developing a numerical model for examining the particle and fluid (air) interaction in a fluidized bed and use it as a tool to investigate biomass particle fluidization in a reactor. The animation on the top right side is the result from our numerical simulation. To validate the numerical model, we have built a cold fluidized bed. The video clip on the left side is taken in our lab, which shows the fluidization of sand particles in a high velocity air flow. The experimentally measured fluid velocity, pressure drop and bed height and the visualized particle distributions will be used to compare with numerical simulation results and to improve our numerical model.

  Particulate Flow and Direct Numerical Simulation

The particulate flow problem has many applications in the fields of chemical, aerospace and environmental engineering as well as in geology and biology, ranging from the transport of radionuclides by sedimentary particles in aquatic environments, to fluidized bed reactors, to droplet formation and combustion. Because of the importance of these applications, the fluid-particles interaction problems have been attracting considerable attention, both experimentally and numerically.

The direct numerical simulation (DNS) is one of the numerical simulation technqiues for particulate flow modeling. DNS accounts for the solid and fluid interaction by solving the Navier-Stokes equations for the fluid phase and the initial value problem for the motion of the particles simultaneously. With the rise of computer power, the DNS method is becoming a more enabling and popular approach to study complex particulate flow problems.

Some recent simulation results using DNS method

A droplet impacts onto a plate under gravity

Evolution of a bubble in a fluidized bed. 12,000 particles are used. Color is used for better viewing of particle mixing.

Rising of two hot particles due to buoyancy force. Color indicates temperature.

  Use of CFD for Industrial Applications

Heavy-Duty Truck Underhood Thermal Analysis (Collaborative Research with Dr. S-L Mao of LANL)To verify and validate the concept design and prototyping of heavy-duty trucks including speed of cooling fan, core sizes of radiator, hydraulic cooler and condenser, we recently conducted a full-size of machine-level underhood CFD thermal analysis using commercial software in combination with user self-defined functions. We investigate the inlet and outlet temperature, pressure drop, surface temperature of machine components, thickness of insulation and design options of heat shield for the new design. Reynolds Averaged Navier-Stokes (RANS) model is employed to simulate underhood turbulent flow and a net-radiation enclosure model is used in the energy equation.


Flow Accelerated Corrosion (Collaborative Research with Drs. Nasrazadani of UNT and Mao of LANL)It is well established that the rate of Flow Accelerated Corrosion (FAC)in pipe is influenced by hydrodynamic factors. The Most severe wall thinning in pipe is observed in the vicinity of welds. We are using CFD to investigate the pressure, the turbulence intensity, the mass transfer rate of the ferrous ions into flows, and the surface shear stress in the region near welds. These simulation results will be able to provide insight on FAC and help us to identify the major factors attributed to FAC.

 Turbulent Modeling and Large Eddy Simulation

Turbulent flow contains eddys of all kinds of scales; it is impossible to explicitly resolve the motion of small eddies, which requires a grid that is finer than the size of small eddies in numerical simulation. However, the small eddies tend to be isotropic and have a universal character, so we can model the effect of small eddies and add these effects into the equations that govern the motion of large eddies, which are the filtered (eddies less than certain size are excluded) Navier-Stokes equations, and solve the equations numerically. This technique is called Large-Eddy Simulation (LES).

LES becomes increasingly a popular computing approach for predictiong turbulent flows. One of my research interests is to develop LES model that is suitable for atmospheric boundary layer flow modeling, which has applications in wind energy related area.

The animation on the right side shows how the concentration of mass or heat propagates from a source due to strong wind. The turbulent flow is modeled by LES method. The contour indicates the concentration of mass or heat.


  People in Multiphase Flow Simulation Laboratory

Principle Investigator

Zhi-Gang Feng
Associate Professor  [CV]
zhigang {dot} feng {at} utsa {dot} edu
+1 210 458 5737
EB 3.04.14


Miguel Cortina
PhD Student
mecortinpo {at} gmail {dot} com
AET 2.206
Joshua Conner
Master Student
joshua {dot} conner {at} my {dot} utsa {dot} com
AET 2.206
Yifei Duan
PhD Student
duanyifeixp {at} gmail {dot} com
mecortinpo {at} gmail {dot} com
AET 2.206
Steven Cooks
Master Student
stcjr89 {at} gmail {dot} com
AET 2.206
Kody Smajstrla
Master Student
ksmajstrla {at} hotmail {dot} com
AET 2.206
Jason Gatewood
Master Student
jason_gatewood {at} msn {dot} com
Cenk Sarikaya
Master Student
zfh258 {at} my {dot} utsa {dot} edu
AET 2.206
Adam Roig
Master Student
adam {dot} roig {at} intertek {dot} com
Joshua Moran
Undergraduate Student
ygz355 {at} my {dot} utsa {dot} edu
AET 2.206
Silvia Murguia
Undergraduate Student
smurguia6 {at} gmail {dot} com
AET 2.206
Carlos Mendez
Undergraduate Student
c {dot} mendez70 {at} yahoo {dot} com
AET 2.206

Former Students

Erwin Garcia
Master Student
Research topic: “Lattice Boltzmann Simulation of Pressure Drop for Laminar Flows in Wavy Pipes.”
Gem Musong Samuel
Master Student
Thesis topic: “A Three Dimensional Immersed Boundary-Based Method for the Free and Combined Convective Heat Transfer from Spherical Bodies.”
Gregory Sloan
Master Student
Thesis topic: “Development and Parallelization of A Direct Numerical Simulation to Study the Formation and Transport of Nanoparticle Clusters in a Viscous Fluid.”
Karim Ebadinia
Master Student
Thesis topic: “Wood Pyrolysis in Fluidized Bed Reactor.”
Ricardo Rodriguez
Undergraduate Student
Eduardo Rodriguez
Undergraduate Student
Maria Andersson
Graduate Student
Eric Stewart
Undergraduate Student
Samuel Yochmowitz
Undergraduate Student
James O’Grady
Master Student
Project topic: “Finite Difference Modeling of Drying 3D Printed Parts.”
Basu Paudel
Master Student
Thesis topic: “Experimental study on biomass and sand mixture in a fluidized bed.”
Charles Obuseh
Master Student
Thesis topic: “Quasi-three dimensional experiments on liquid-solid fluidized bed of three different particls in two different distributors.”


  Recent Publications

    A complete list of publications and their citations can be found on Google Scholar

 Journal Papers Published between 2009-2015

  • Feng, Z-G, Alatawi, E. S., Roig, A., and Sarikaya, C.(2015), “A resolved Eulerian-Lagrangian simulation of fluidization of 1204 heated spheres in a bed with heat transfer.” ASME Journal of Fluid Engineering, accepted.
  • Musong, S. and Feng, Z-G (2014), “Mixed convective heat transfer from a heated sphere at an arbitrary incident flow angle in laminar flows,” International Journal of Heat and Mass Transfer, 78:34-44.
  • Feng, Z-G and Musong, S. (2014), “Direct numerical simulation of heat and mass transfer of spheres in a fluidized bed.” Powder Technology, 262:62-70.
  • Feng, Z-G (2014),”Direct numerical simulation of forced convective heat transfer from a heated rotating sphere in laminar flows,” Journal of Heat Transfer, 136:041707.
  • Feng, Z-G, Ponton, M. E., Michaelides, E. E., and Mao, S-L (2013), “Using the direct numerical simulation to compute the slip boundary condition of the solid phase in two-fluid model simulations.” Powder Tech., 10.1016/j.powtec.2014.01.020.
  • Paudel, B., and Feng, Z-G (2013),”Prediction of minimum fluidization velocity for binary mixtures of biomass and inert particles,” Powder Technology, 237:134-140.
  • Feng, Z-G (2013),”Forced heat and mass transfer from a slightly deformed sphere at small but finite Peclet numbers in Stokes flow,” Journal of Heat Transfer, 135:081702.
  • Obuseh, C.C., Feng, Z-G, Paudel, B.D.(2012),”An Experimental Study on Fluidization of Binary Mixture in Particulate Flows,” Journal of Dispersion Science and Technology 33 (9): 1379-1384.
  • Feng, Z-G, and Michaelides, E.E. (2012), “Heat transfer from a nano-sphere with temperature and velocity discontinuities at the interface,” International Journal of Heat and Mass Transfer, 55:6491-6498.
  • Davis, A.P., Michaelides, E.E., Feng, Z-G (2012),”Particle velocity near vertical boundaries: A source of uncertainty in two-fluid models,” Powder Technology 220:15-23.
  • Feng, Z-G, Michaelides, E.E., and Mao, S.(2012),”On the drag force of a viscous sphere with interfacial slip at small but finite Reynolds numbers.” Fluid Dynamics Research, 44 (2):0255.
  • Redrow, J, Mao, S-L, Celik,I, Posada, J.A.,and Feng, Z-G (2011),”Modeling the evaporation and dispersion of airborne sputum droplets expelled from a human cough,” J. Building and Environment, 46(10):2042-2051.
  • Yang, B.J., Mao, S., Altin, O.,Feng, Z-G. (2011),”Condensation Analysis of Exhaust Gas Recirculation System for Heavy-Duty Trucks,” Journal of Thermal Science and Engineering Applications, 3(4):#041007.
  • Mao, S-L, Feng, Z-G, Michaelides, E.E.(2010), “Off-highway heavy-duty truck under-hood thermal analysis,” Applied Thermal Engineering, 30 (13), 1726-1733.
  • Feng, Z-G (2010),”A correlation of the drag force coefficient on a sphere with interface slip at low and intermediate Reynolds numbers.” Journal Dispersion Science Technology, 31:968-974
  • Feng, Z-G, Michaelides, E.E., and Mao, S-L(2010),”A three-dimensional resolved discrete particle method for studying particle wall collision in a viscous fluid,” ASME J. Fluids Engineering, 132 (9), #091302.
  • Feng, Z-G, and Michaelides, E.E. (2009), “Secondary flow within a river bed and contaminant transport.” Environmental Fluid Mechanics, DOI 10.1007/s10652-009-9132-9
  • Feng, Z-G, and Michaelides, E.E. (2009), “Heat transfer in particulate flows with Direct Numerical Simulation (DNS).” International Journal of Heat Mass Transfer, 52:777-786
  • Feng, Z-G, and Michaelides, E. E. (2009), “Robust treatment of no-slip boundary condition and velocity updating for the Lattice-Boltzmann Simulation of Particulate Flows.” Computers and Fluids, 38:370-381

 Conference Papers and Presentations between 2009-2014

  • Smajstrla, K and Feng, Z-G (2014), “A one-dimensional model for gas-solid heat transfer in pneumatic conveying,” ASME 2014 International Mechanical Engineering Congress and Exposition, November 14-20, 2014, Montreal, Canada
  • Feng, Z-G and Ponton, M. E., “Smoothed Particle Hydrodynamics (SPH) method for studying heat and mass transfer between fluid and solid.” ASME 2014 International Mechanical Engineering Congress and Exposition, November 14-20, 2014, Montreal, Canada
  • Yao, Z., Feng, Z., Qin, Z., & Chen, Z., “Heat transfer enhancement for turbulent flows in corrugated tubes,” ASME 2014 International Mechanical Engineering Congress and Exposition, November 14-20, 2014, Montreal, Canada.
  • Musong, S, Feng, Z-G, K. Chen, and Xu, Q-W (2014), “Effects of Rod Shapes on the Drag Force of Particles in a Shear ASME 2014 International Mechanical Engineering Congress and Exposition, November 14-20, 2014, Montreal, Canada
  • Feng, Z-G and Roig, A, “Direct numerical simulation of particle heat and mass transfer in a fluidized bed.” ASME Gas-Solid Symposium, 2014, Chicago.
  • Feng, Z-G, Musong, S., and Michaelides, E. E., “A three dimensional immersed boundary method for free convection from single spheres and aggregates.” ASME Gas-Solid Symposium, 2014, Chicago.
  • Musong, S., Cortina, M., Feng, Z-G, “A three dimensional direct numerical simulation method for solving heat transfer of particulate flows.” 2013 NETL Workshop on Multiphase Flow Science, Morgantown.
  • Feng, Z-G, G. Sloan, K. Bhaganagar, and D. Banerjee, “Numerical simulation of nanoparticle clustering with experimental validation,” APS Division of Fluid Dynamics, November 18-20, 2012; San Diego, California.
  • Feng, Z-G, and M. Andersson, “Modeling flows in porous media using immersed boundary based lattice Boltzmann method,” ASME International Mechanical Engineering Congress & Exposition, Nov.10-15, Houston, TX; Paper Number: IMECE2012-89427.
  • Sloan, G., Feng, Z-G, K. Bhaganagar, K, and Banerjee, D., “Coupled direct numerical simulation and experimental approach to develop framework for nanofluids,” ASME International Mechanical Engineering Congress & Exposition, Nov.10-15, Houston, TX. Paper number: IMECE2012-89271.
  • Cortina, M., and Feng, Z-G, “A comparable study on particle sedimentation by the resolved discrete particle method and two-fluid model,” 2012 San Antonio Simulation and Visualization Symposium, Nov. 12-13, 2012.
  • Feng, Z-G, Feng, Y-S, and M. Andersson, “Shape effects on the drag force and motion of nano and micro particles in low Reynolds number flows,” ASME International Mechanical Engineering Congress & Exposition, Nov.10-15, Houston, TX. Paper number: IMECE2012-89469.
  • Feng, Z-G, Musong, S., Sloan, G., Anderson, M., and Stewart, E., “The effect of neighboring particles on the dynamics of a particle settling in a viscous fluid,” DOE/NETL Conference on Multiphase Flow Science, May 22-24, 2012, Morgantown, WV.
  • Musong, S. and Feng, Z-G, “The effect of model parameters of the soft-sphere scheme on particle-particle collisions,” DOE/NETL Conference on Multiphase Flow Science, May 22-24, 2012, Morgantown, WV.
  • Feng, Z-G, Sloan, G., Musong S., Davis, A. D., Ebadinia, K., Cook, S., “Use of a DNS method to reduce uncertainties in two-fluid models,” 2012 University Coal Research/HBCU and Other Minority Institutions Contractors Review Conference, May 30-31, 2012, Pittsburg, PA.
  • Feng, Z-G., Michaelides, E. E., and Mao, S, “A multilevel simulation approach to derive the slip boundary condition of the solid phase in two-fluid models,” 64th Annual Meeting of the APS Division of Fluid Dynamics, November 20-22, 2011; Baltimore, Maryland.
  • Z-G Feng, “Computational Modeling of Biological Flows,” Navy Tri-Service Research Laboratory, November 4, 2011.
  • B. Paudel and Z-G Feng, “Fluidization of inert, biomass particles and biomass/sand mixtures,” 4th ASNEngr Annual Conference and Meeting and 4th CAN-USA Annual Development Conference. July 09-10, 2011. Houston, Texas.
  • Feng, Z-G, et al., “Use of an Accurate DNS Particulate Flow Method to Study Boundary Conditions of the Solid Phase in Two-Fluid Model,” the 2011 University Coal Research/HBCU Conference, June 7-8, 2011. Pittsburgh, Pennsylvania.
  • Davis, A. P., Michaelides, E. E., and Feng, Z-G, “Particle velocity near vertical boundary a source of uncertainty in two-fluid models,” 7th International Conference on Multiphase Flow, FL. May 30-June 4, 2010.
  • Feng, Z-G and Michaelides, E.E, “Application of the Immersed boundary method and direct numerical simulation for the heat transfer from particles,” ASME Fluids Engineering Division Summer Conference, Vail, Colorado, August 2009.

 Book and Book Chapter

  • Michaelides, E. E. and Feng, Z-G (2015). “Chapter 1: Basic concepts and definitions.” Multiphase Flow Handbook -2nd edition (in press).
  • Feng, Z-G and Michaelides, E. E. (2015). “Chapter 3: The immersed boundary method to study interactions of particulate flows with heat transfer.” Multiphase Flow Handbook -2nd edition (in press).
  • Michaelides, E. E. and Feng, Z-G (2015). “Implementation of the immersed boundary method to study interactions of fluids with particles, bubbles and drops.” Progress in Colloid and Interface Science, Vol.7. Taylor & Francis (in press).
  • Tran-Cong, S. and Feng, Z-G (editors) (2001). “Book of abstracts of the 4th International Conference on Multiphase Flow.” New Orleans, Louisiana, USA, May 27-June 1, 2001.