Luoding Zhu

An osteocyte is a bone cell situated inside a hard bone matrix in an interstice (lacuna). It has many dendritic structures called cellular processes that radiate outward from the cell through the bone matrix via cylindrical openings (canaliculi). Osteocytes can sense stress and strain applied by the interstitial fluid flow and respond by releasing biochemical signals that regulate bone remodeling. In vitro experiments have suggested that the stress and strain typically experienced at the macroscale tissue level have to be amplified 10× in order for osteocytes to have a significant response in vivo. This stress and strain amplification mechanism is not yet well understood. Previous studies suggest that the processes are the primary sites for mechanosensation thanks to the tethering elements that attach the process membrane to the canalicular wall. However, there are other potential factors which may also contribute to stress and strain amplification, such as canalicular wall geometry and osteocyte-associated proteins in the interstitial space called pericellular matrix. In this work, we perform computational studies to study how canalicular wall roughness affects stress and strain amplification. Our major finding is that the wall roughness induces significantly greater wall shear stress (WSS) on the process when the wall roughness increases flow resistance; and the roughness has relatively smaller influence on the WSS when the resistance remains the same.

Osteocytes are bone cells that can sense mechanical cues (stress and strain) and respond by releasing biochemical signals that direct bone remodeling. This process is called mechanotransduction which, in osteocytes, is not well understood yet because in vivo studies have proven difficult due to the complexity and inaccessibility of the flow–osteocyte lacuna-canaliculi system. While in silico studies (modeling and simulation) have become powerful, currently computational studies for the system often omit the fluid–structure interaction (FSI) between the cell and the surrounding fluids. To investigate the role of FSI in osteocyte mechanotransduction, we introduce a two-dimensional coarse-grained yet integrative model for flow–osteocyte interaction in a lacuno-canalicular network. The model uses the lattice Boltzmann immersed boundary framework to incorporate the flexible osteocyte (membrane, cytoskeleton, and cytosol), its processes, the interstitial fluid, and the rigid extracellular matrix that encases the system. One major result of our model is that the stress and strain tend to attain their local maxima near the regions where the processes meet the membrane of the main body.

Experiments have shown that external mechanical loading plays an important role in bone development and remodeling. In fact, recent research has provided evidence that osteocytes can sense such loading and respond by releasing biochemical signals (mechanotransduction, MT) that initiate bone degradation or growth. Many aspects on MT remain unclear, especially at the cellular level. Because of the extreme hardness of the bone matrix and complexity of the microenvironment that an osteocyte lives in, in vivo studies are difficult; in contrast, modeling and simulation are viable approaches. Although many computational studies have been carried out, the complex geometry that can involve 60+ irregular canaliculi is often simplified to a select few straight tubes or channels. In addition, the pericellular matrix (PCM) is usually not considered. To better understand the effects of these frequently neglected aspects, we use the lattice Boltzmann equations to model the fluid flow over an osteocyte in a lacuno-canalicular network in two dimensions. We focus on the influences of the number/geometry of the canaliculi and the effects of the PCM on the fluid wall shear stress (WSS) and normal stress (WNS) on an osteocyte surface. We consider 16, 32, and 64 canaliculi using one randomly generated geometry for each of the 16 and 32 canaliculi cases and three geometries for the 64 canaliculi case. We also consider 0%, 5%, 10%, 20%, and 40% pericellular matrix density. Numerical results on the WSS and WNS distributions and on the velocity field are visualized, compared, and analyzed. Our major results are as follows: (1) the fluid flow generates significantly greater force on the surface of the osteocyte if the model includes the pericellular matrix (PCM); (2) in the absence of PCM, the average magnitudes of the stresses on the osteocyte surface are not significantly altered by the number and geometry of the canaliculi despite some quantitative influence of the latter on overall variation and distribution of those stresses; and (3) the dimensionless stress (stress after non-dimensionalization) on the osteocyte surface scales approximately as the reciprocal of the Reynolds number and increasing PCM density in the canaliculi reduces the range of Reynolds number values for which the scaling law holds.

Late-stage kidney disease patients have to rely on hemodialysis for the maintenance of their regular lives. Arteriovenous graft (AVG) is one of the commonly used devices for dialysis. However, this artificially created shunt may get clotted and eventually causes the dialysis to fail. The culprit behind the AVG clotting and failure is the intimal hyperplasia (IH), the gradual thickening of vein-wall in the vicinity of the blood vessel-graft conjunctions. The mechanism of IH is not well understood despite extensive studies. In this work, we investigate the effects of the IH development, including its location and severity on the flow and force fields in the distal AVG anastomosis using computational fluid dynamics. The stenosis due to IH is modeled in the shape of a Gaussian function with two free parameters. The blood is modeled as a viscous incompressible fluid, and the blood flow (pulsatile) is governed by the Navier–Stokes equations which are numerically solved by the lattice Boltzmann model (D3Q19). The fluid-structure interaction is modeled by the immersed boundary framework. Our computational results show that the IH severity has the most significant influences on the wall shear stress, wall-normal stress, and the axial oscillating index. The stenosis location and flow pulsatility do not have pronounced effects on flow and force fields. Our results indicate that the IH progression tends to exacerbate the disease and accelerate the closure of the vein lumen, and hence the dialysis failure.

This work is motivated by an experiment of microbubble transport in a polymer microfluidic gas generation device where coalescence-induced detachment exhibits. We numerically study three-dimensional microbubble coalescence using the graphics processing unit accelerating free energy lattice Boltzmann method with cubic polynomial boundary conditions. The focus is on the coalescence-induced microbubble detachment (CIMD) in microfluidics. From the experimental observation, we identified that size inequality between two-parent bubbles and the size of the father (large) bubble are key factors to determine if a CIMD will occur. First, the analytical relationship between equilibrium contact angle and dimensionless wetting potential and experimental results of coalescence with and without CIMD are employed for the verification and validation, respectively. From eighteen experimental and computational cases, we derive a new criterion for CIMD: CIMD occurs when the two-parent bubbles are (nearly) equal with a relatively large radius. The underlying mechanism behind this criterion is explored by the time evolution of the velocity vector field, vorticity field, and kinetic energy in the entire coalescence. It is found that the symmetric capillary force drives the formation of vertical flow stream to the horizontal alignment of parent bubbles and the blockage of the downward stream due to the solid interface promotes the intensity of the upward stream. Meanwhile, large-sized parent bubbles transfer a large amount of kinetic energy from the initial free surface energy, which is essential to lead a CIMD in the post-coalescence stage. Such a new criterion is expected to impact the design and optimization of microfluidics in various applications.

We consider a deformable plate interacting with a non-Newtonian fluid flow in three dimensions as a simple model problem for fluid-structure-interaction phenomena in life sciences (e.g., red blood cell interacting with blood flow). A power-law function is used for the constitutive equation of the non-Newtonian fluid. The lattice Boltzmann equation (the D3Q19 model) is used for modeling the fluid flow. The immersed boundary (IB) method is used for modeling the flexible plate and handling the fluid-plate interaction. The plate drag and its scaling are studied; the influences of three dimensionless parameters (power-law exponent, bending modulus, and generalized Reynolds number) are investigated.