In this study, the new convection heat transfer concept of sweeping convection, using large particles flowing with a fluid in a channel, is considered. This novel concept is inspired by the gas exchange process in alveolar capillaries, where red blood cells (RBCs) flow with blood plasma, yielding very high gas transfer efficiency. An important characteristic of alveolar capillary blood flow, believed to be related to the high efficiency of the lungs, is the snug fitting of the RBCs into the capillaries. This tight fitting sets the RBCs (particles) acting like pistons as they flow downstream with the plasma (fluid), facilitating mass transfer by disrupting the velocity and concentration boundary layers that otherwise develop along the channel. The analogy between mass and heat transfer supports the expectation this effect would be presented also under heat convection. Hence, the sweeping of the boundary layers by solid particles, and the resulting increased mixing, is expected to yield higher convection heat transfer coefficients, as compared to the heat transfer coefficient achieved when a fluid clear of particles is flowing through a heated (or cooled) channel. Therefore, results of a computational and analytical investigation are presented here, followed by some experimental testing of sweeping heat convection.

The primary system considered here consists of a regular, straight channel through which a Newtonian liquid (water, with uniform and constant properties) flows with or without spherical solid particles (constant and uniform properties) immersed in it. Simulations were performed for the fluid-particle flow beginning from a quiescent state (named “startup flow”) through a heated channel, either with a constant and uniform surface temperature or with a constant and uniform surface heat flux condition. An additional group of simulations were performed considering the flow to begin from a fully developed clear fluid state in which particles were included starting to flow in the channel prior to the heated section. The simulations took place under the laminar flow regime (27 < Re < 564). Two Fluid Structure Interaction (FSI) methods were utilized to simulate the flowing of the particle with the fluid, namely: Moving Mesh and Immersed Solid. The numerical model and the corresponding CFD code were developed using ANSYS CFX software as basis. The results reveal the particle effect on the convection process to be localized and yet very significant, with the surface-averaged Nu number increasing by up to 60% at a specific time in the process and up to 35% when time averaged during the entire process; and with negligible pressure drop effect across the channel. A focused analysis of the flow at the gap between the moving particle and the channel surfaces yielded an analytical prediction of the fluid and particle velocities, which was validated by the numerical results. The results also show the three essential effects of the particle as it sweeps the fluid in the channel, namely: stretching the flow upstream (dragging additional cold fluid into the channel), compressing it downstream (pushing hot fluid out of the channel), and accelerating it through the gap (stronger surface convection). The somewhat unexpected result of high heat transfer coefficient with negligible pressure drop increase reinforces the expectation of sweeping convection being a very effective new convection mechanism with tremendous practical potential, particularly for its inherently low pressure drop penalty.

In the experimental effort of this work, a new particulate flow circulation system is introduced, not only to provide a tool to verify numerical simulation and analytical results but also to study the practical aspects of implementing sweeping convection in a convection heat transfer system. The novel experimental apparatus, designed as a flow loop, uses a vortex-based pumping effect created by an impeller to set the fluid (liquid) and particles in motion inside a container that then feeds the flow through a tube linked to a heated pipe for testing. The vortex-based pumping system performs very well, alleviating three major draw backs of fluid-particle flows, namely pumping without damaging the particles, and particle agglomeration (clogging) and settling in the flow system. For the heat transfer investigation, the testing tube section was fitted with a uniform surface heater and instrumented with several thermocouples for fluid and surface temperature monitoring. Two types of experimental procedures were then performed: (1) a Local Effect Study (LES), which focuses on a single or only a few particles circulating in the test section at a time; and, (2) the Overall Effect Study (OES), which investigates the effect of a large number of particles placed in the circulation system. Results from the LES procedure shows the same flow and heat transfer characteristics as observed in the numerical simulations, particularly the local channel surface temperature drop and recovery upon the passing of a particle. The OES procedure results indicated the average surface temperature of the locations near the outlet in particle flow to be significantly lower than the temperature achieved under clear flow for the same flow conditions. The results of the conducted experimental tests support qualitatively the findings of the numerical simulations, and provide venues for expanding this study.

KEY WORDS: Sweeping convection, Convective heat transfer, Particle flow

Degree Date

Fall 2020

Document Type


Degree Name



Mechanical Engineering


Dr. José L. Lage

Subject Area

Mechanical Engineering



I would express my sincere gratitude to my honorable advisor, Prof. José Lage, for his continuous guidance, encouragement and precious advice throughout my research. His insights and suggestions contributed to the progress of my research from beginning to end.

Many thanks to Prof. Ali Beskok, Dr. David Willis, Dr. Usama El Shamy and Dr. Silvio Junqueira for their efforts and valuable comments as members of my Ph.D. advisory committee. It has been an honor for me to interact with them in such capacity. I am greatly thankful to all of my friends at SMU for their support. Last not by least, I am very grateful for the loving support of my beloved family. This work has been financially supported by NSF Grant No. CBET-1404017. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation.

Number of Pages




Creative Commons License

Creative Commons Attribution-Noncommercial 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License