Publications

Carbon nanomaterials, such as carbon nanotubes (CNTs) and carbon nanofibers (CNFs), are chemically inert in their highly graphitic forms. Various post processing methods can activate their surfaces to enhance their interactions with a host matrix in a nanocomposite. Chemical surface functionalization is used often. This method however can lead to major strength loss in nanomaterials stemming from induced surface defects (changing sp² bonds to sp³ bonds). In this manuscript, we have experimentally studied the mechanical properties of the individual, pyrolysis-fabricated CNFs. These CNFs have a highly crosslinked 3D network of C–C bonds. The strength of CNFs has been studied as a function of O/C ratio. The loss in strength due to functionalization has been compared to that of other carbon nanomaterials with layered strcutures (CNT and graphene). Comparisons were also made with carbon microfibers. Fracture strength estimations of the critical flaw size in CNFs, CNTs and graphene were also made. The results revealed that despite having high surface area, carbon nanomaterials with crosslinked microstructure are resilient to flaws as big (deep) as 10–30 nm, while nanomaterials with layered structure (such as CNTs) experience a dramatic loss in strength with much lower flaw sizes. Hence, it seems that graphitic nanomaterials such as graphene and CNT have high strength that, although higher than CNFs, comes at a cost to flaw tolerance and robustness. Since failure is often progressive, this work demonstrates a benefit that crosslinked nanomaterials have over highly graphitic ones, such as CNTs, in load bearing applications.

Crosslinked network microstrucutre of carbon nanomaterials promotes flaw-tolerant mechanical response

This paper presents a numerical study for nonlinear rotordynamic response with bifurcations of tilting pad journal bearings when pad–pivot friction forces are taken into account. A Stribeck friction model is employed to determine the friction coefficient for the contacts between the pads and the spherical-type pivots. The boundary/mixed/hydrodynamic friction mode is determined for each pad surface based on the instantaneous angular motion of the pads. A Jeffcott type rotor supported on 5-pad tilting pad journal bearings is used for the structural model, and finite element fluid film models are utilized to calculate the reaction forces and moments on the pads. The simulation results show that pad–pivot friction plays an important role in determining the stability of the rotor system. For the autonomous condition, the friction induces a Hopf bifurcation and generates limit cycles at high rotor spin speed (>14 krpm), which were originally stable equilibrium states with a no friction condition. For the nonautonomous condition, the 1× synchronous response becomes subsynchronous/quasiperiodic responses in the high-speed range (>14 krpm) with the appearances of Neimark-Sacker bifurcations. It is shown that the outbreak points and corresponding response types are highly dependent on the state of disk imbalance. A comparison of the linear and nonlinear models clearly illustrates the importance of retaining nonlinear forces to determine potential deleterious vibration.

Pad-Pivot Friction Effect on Nonlinear Response of a Rotor Supported by Tilting-Pad Journal Bearings

The authors present an improved formulation for the axisymmetric solid harmonic finite element (FE) modeling of a flexible, spinning rotor. A thorough comparison of beam-type FE and axisymmetric solid FE rotor models is presented, indicating the errors that result from beam FE usage for various nondimensional rotor topologies. The axisymmetric rotor is meshed in only two dimensions: axial and radial, with both displacement fields being represented with Fourier series expansions. Centrifugal stress-stiffening and spin-softening effects are included in all elements and most importantly in modeling flexible disks. Beam FE and axisymmetric FE natural frequencies, mode shapes, and critical speeds are compared to identify shaft geometries where the beam model yields a significant error. Finally, limitations of beam FE models and guidance for utilizing axisymmetric solid FE models in rotor dynamic simulations are provided.

An Enhanced Axisymmetric Solid Element for Rotor Dynamic Model Improvement

Modern high performance turbomachines frequently operate in supercritical condition above their first critical speed, rendering these machines prone to rotordynamic instability. The American Petroleum Institute (API) standards require advanced simulation models for level II stability analysis of impellers. Such data are then incorporated into rotor-bearing vibration response models. Despite recent advancements in high fidelity, general modeling (i.e., three-dimensional viscous transient nonaxisymmetric model) of closed impeller rotordynamic forces, no such general model is available for open impellers, especially the centrifugal type. This paper extends the transient computational fluid dynamics (CFD) model previously used for closed impellers to open impellers. The recent model uses a phase modulated, multifrequency approach for enhanced computational efficiency and robustness. Results are validated against literature experiments at design and off-flow condition. The model is further applied to a spectrum of specific speeds to extract the dimensionless rotordynamic forces for each class of impellers at design and off-flow conditions. Such dimensionless force data can be used to estimate the rotordynamic forces of impellers with similar specific speed. Depending on specific speed and the relative flow coefficient, many of these impellers are found to be excited by forward or backward whirl. Strong interaction with rotating stall typically appears in the force data at off-flow condition. Simulations of the isolated leakage path model (ILPM) for equivalent closed impellers reveal similar bumps and dips associated with highly swirling inflow which naturally occurs at part flow condition.

A Transient Computational Fluid Dynamics, Phase Modulated, Multi-Frequency Approach for Impeller Rotordynamic Forces

This paper presents the first simulation model of a tilting pad journal bearing (TPJB) using three-dimensional (3D) computational fluid dynamics (CFD), including multiphase flow, thermal-fluid, transitional turbulence, and thermal deformation of the shaft and pads employing two-way fluid–structure interaction (FSI). Part I presents a modeling method for the static performance. The model includes flow between pads BP, which eliminates the use of an uncertain, mixing coefficient (MC) in Reynold’s equation approaches. The CFD model is benchmarked with Reynold’s model with a 3D thermal-film, when the CFD model boundary conditions are consistent with the Reynolds boundary conditions. The Reynolds model employs an oversimplified MC representation of the three-dimensional mixing effect of the BP flow and heat transfer, and it also employs simplifying assumptions for the flow and heat transfer within the thin film between the journal and bearing. This manufactured comparison shows good agreement between the CFD and Reynold’s equation models. The CFD model is generalized by removing these fictitious boundary conditions on pad inlets and outlets and instead models the flow and temperature between pads. The results show that Reynold’s model MC approach can lead to significant differences with the CFD model including detailed flow and thermal modeling between pads. Thus, the CFD approach provides increased reliability of predictions. The paper provides an instructive methodology including detailed steps for properly applying CFD to tilt pad bearing modeling. Parts I and II focus on predicting static and dynamic response characteristic responses, respectively.

3D Thermo-Elasto-Hydrodynamic CFD Model of a Tilting Pad Journal Bearing – Part II Dynamic Response

This paper presents the first simulation model of a tilting pad journal bearing (TPJB) using three-dimensional (3D) computational fluid dynamics (CFD), including multiphase flow, thermal-fluid, transitional turbulence, and thermal deformation of the shaft and pads employing two-way fluid–structure interaction (FSI). Part I presents a modeling method for the static performance. The model includes flow between pads BP, which eliminates the use of an uncertain, mixing coefficient (MC) in Reynold’s equation approaches. The CFD model is benchmarked with Reynold’s model with a 3D thermal-film, when the CFD model boundary conditions are consistent with the Reynolds boundary conditions. The Reynolds model employs an oversimplified MC representation of the three-dimensional mixing effect of the BP flow and heat transfer, and it also employs simplifying assumptions for the flow and heat transfer within the thin film between the journal and bearing. This manufactured comparison shows good agreement between the CFD and Reynold’s equation models. The CFD model is generalized by removing these fictitious boundary conditions on pad inlets and outlets and instead models the flow and temperature between pads. The results show that Reynold’s model MC approach can lead to significant differences with the CFD model including detailed flow and thermal modeling between pads. Thus, the CFD approach provides increased reliability of predictions. The paper provides an instructive methodology including detailed steps for properly applying CFD to tilt pad bearing modeling. Parts I and II focus on predicting static and dynamic response characteristic responses, respectively.

3D Thermo-Elasto-Hydrodynamic CFD Model of a Tilting Pad Journal Bearing – Part I Static Response

Experiments and Computational Fluid Dynamics (CFD) simulations were performed to investigate the colloidal fouling control of a vibration enhanced reverse osmosis (VERO) membrane system for up to 12 h of operation time. A porous cake mass transfer model for the reverse osmosis (RO) colloidal fouling analysis was derived by combining the cake filtration model, cake enhanced osmotic pressure (CEOP) effect, the critical flux theory, and the solute convection-diffusion model. The process of colloidal particle deposition and fouling formation on the membrane surface were visualized using this model. The permeate flux variation and cake layer distribution along the membrane were depicted. Systematic studies including different initial permeate fluxes, Reynolds numbers, particle concentrations, and vibration frequencies were carried out to investigate the system performance under different operation conditions. The results suggest that colloidal fouling induced permeate flux decline could be improved by the high-frequency vibration in the VERO module. Both simulations and experiments demonstrated that, with a fixed vibration amplitude, membrane module with higher vibration frequencies will have less NaCl accumulation and higher permeate flux under the influence of colloidal fouling.

Permeate Flux Increase by Colloidal Fouling Control in a Vibration Enhanced Reverse Osmosis Membrane Desalination System

Concentration polarization and surface fouling may be two of the most remarkable features in the pressure-driven membrane filtration process. A new numerical simulation model is proposed in this paper to study the concentration polarization (CP) and the inorganic fouling growth. The numerical study is based on the lattice Boltzmann method (LBM), which allows a simultaneous solution of the Navier–Stokes equations and the convection-diffusion equation. Simulation results are verified by comparisons with published CP and permeate flux data under the same operating conditions. Then the model is extended to predict CP in a spacer filled desalination channel. The prediction result indicates that there is a higher fouling potential near the spacer filaments due to higher CP values in that area. Coupling of the CP prediction model with gypsum growth kinetics provides an approach to study the inorganic fouling growth on the membrane surface at a single crystal level, with respect to a given solution supersaturation near the membrane surface. Predicted equivalent radius and accumulated mass of the growing gypsum crystal, under the effects of growth retardation by bicarbonate, agree with published test data and analytical results. The presented numerical model enables a direct evaluation of the impacts on the surface crystal development in the presence of antiscalants. This numerical model can be applied to identify suitable operating conditions, assist in dose selection of antiscalants when required properties are available, and predict the fouling mitigation effects in the pressure-driven membrane filtration process.

Numerical Modeling of Concentration Polarization and Inorganic Fouling Growth in the Pressure-Driven Membrane Filtration Process