Symposium‎ > ‎

2014 ME Abstracts

Large Eddy/Filtered Mass Density Function Simulation Of Colorless Distributed Combustion

Authors: Husam Abdulrahman; Farhad Jaberi

Abstract: Development of efficient and low-emission colorless distributed combustion

(CDC) systems for gas turbine applications require careful examination of the role of various flow and combustion parameters. Numerical simulations of CDC in a laboratory-scale combustor have been conducted to carefully examine the effects of these parameters on the CDC. The computational model is based on a hybrid modeling approach combining large eddy simulation (LES) and filtered mass density function (FMDF) and high order numerical methods with complex chemical kinetics. The simulated combustor operates based on the principle of high temperature air combustion (HiTAC) and has shown to significantly reduce the NOx, and CO emissions while improving the reaction pattern factor and stability without using any flame stabilizer and with low pressure drop and noise. The focus of the current work is to investigate the mixing of air and hydrocarbon fuels, the non-premixed and premixed reactions and emission levels within the combustor by the LES/FMDF with the reduced chemical kinetic mechanisms for the same flow conditions and configurations investigated experimentally. The main goal is to develop better CDC with higher mixing and efficiency, ultra-low emission levels and optimum residence time. The computational results establish the consistency and reliability of the LES/FMDF model and its Lagrangian-Eulerian numerical methodology.


Robust Design Optimization Of Artificial Lateral Line System

Authors: Ali Ahrari; Hong Lei; Kalyanmoy Deb; Xiaobo Tan

Abstract: As an important flow-sensing organ, the lateral line system is involved in various behaviors of fish such as schooling, station holding and object detection. Theoretical work is also conducted on flow modeling to help explain hydrodynamic imaging by fish and extract information in artificial lateral lines. Much of the aforementioned research on information processing for biological and artificial lateral lines has been focused on the localization of a dipole source, which emulates the rhythmic movement of fish body and fins, and has been commonly used as a biological stimulus such as conspecific, predator or prey. Dipole source localization has also played an important role in the development of artificial lateral lines, detection and estimation of nearby fish-like robots, and coordination and control of underwater robot. Despite much research on dipole source localization, there are few studies on optimal design of artificial lateral line system in order to maximize reliability of localization. Presence of diverse sources of uncertainty challenges reliability of localization and practicability of the underlying idea. In this study, different sources of uncertainties are identified and modeled in the problem formulation. A bi-level robust optimization method is proposed to handle inherent uncertainties in the problem. The best design obtained by the proposed optimization method is compared with the initially proposed design on a large number of data set which demonstrates significant advantages of robust optimization over sole engineering intuition. The proposed method clearly illustrates the use of evolutionary algorithms in handling uncertainties and is extendable to similar other problems.

This work was supported in part by BEACON.


Topology Optimization For 3D Printing

Authors: Atra Akandeh; Alejandro Diaz

Abstract: 3D printing (3DP) is a manufacturing process that is well-suited for topology optimization because of the design freedom made possible by 3DP. Most of the work in topology optimization has focused on maximum stiffness design of structures subjected to externally applied mechanical loads.  However, because of thermal gradients, in 3DP the pattern of layout of material may contribute to thermal stresses that affect performance. Traditional topology optimization does not account for this.

In 3DP process-induced stresses may be high enough to affect the structural integrity of the part, and the part will warp.  The aim of this study is to determine the residual stresses during the manufacturing process; identify process conditions that result in reduced residual deformation; and then take thermal stresses into account within a topology optimization formulation.

To simulate the forming mechanism in 3DP, a time-dependent finite element multiphysics model is used. The model simulates a high intensity laser energy source moving along a pre-defined time dependent path, used to solidify the powder being laid down along the path. The model includes the effect of parameters such as convection coefficient, laser velocity, part dimensions, and path direction. It is found that by adjusting the speed of the laser along its path, path direction, and layer thickness, the resulting residual deformation can be reduced or eliminated.


Real-Time Control Of The Boundary Layer Disturbance Induced By A Dynamic Isolated Roughness Element Using Plasma Actuators

Authors: Kyle Bade; Ahmed Naguib

Abstract: Fluid flows found in most engineering applications are generally turbulent (i.e. chaotic, three-dimensional, and time-dependent). In many of these applications, it is desirable to maintain the flow in a laminar (i.e. orderly) state in order to reduce the friction drag between the fluid and solid surfaces. Motivated by the delay/prevention of transition from the laminar to turbulent state, this study examines the ability to sense unsteady disturbances in a Blasius laminar boundary layer and to attenuate their transient growth using plasma actuators. It is well established that a certain ubiquitous type of boundary layer transition (known as “bypass”) is initiated by the formation and growth of unsteady disturbances, known as streaks. Thus, by actively controlling these disturbances in engineering applications, it could be possible to delay transition to turbulence, increasing system efficacy and energy efficiency. In this work, the unsteady streaks are introduced into the boundary layer using an isolated roughness element that is dynamically actuated from flush with the wall to a specified height; resulting in a time varying disturbance. A real-time, closed-loop, feedforward-feedback control system is designed to apply an appropriate voltage to a plasma actuator in order to reduce the roughness induced disturbance. The control system inputs come from two in-wall hot-wire shear stress sensors located within a high-speed streak disturbance, one upstream and one downstream of the plasma actuator. The controller is shown to effectively drive the shear stress at the feedback sensor toward the Blasius level, and effectively reduce the disturbance strength.

This work was supported in part by National Science Foundation (NSF Grant: CMMI 0932546)


Controlled Diffusion Blade: Fluid Mechanical Mechanistic Effects On Fan Performance

Authors: David Barrent; John Foss

Abstract: Experimental work to identify the mechanistic effects on fan performance will be carried out in the Axial Fan Research and Development Facility at the Turbulent Shear Flows Laboratory. Fan performance curves have been generated for both the three and nine blade conditions for the Rotating Controlled Diffusion Blade (RCDB) fan. Calibration of the mass flow sensing device (Morris) inside the AFRD was required to obtain these data. The annular opening for the fan was selected as the inlet for the calibration. The discharge coefficient of this opening was unknown and required for the mass flow calibration. Hotwire surveys along four equally spaced radial lines were used to find the Reynolds number dependent discharge coefficient. Once the calibration for mass flow measurements was complete, performance data for the three and nine blade fan were obtained. Each fan configuration was tested at varying rotational velocities to see if the behavior of the fan would scale in a non-dimensional fashion. The RCDB has instrumented blades with eighteen pressure taps at five radial locations. These will be utilized to find the pressure coefficients along the suction and pressure side of the blade at varying radial locations. Near wake stationary hotwire measurements will be made to characterize the wake of the RCDB.

(Morris, Neal, Foss, and Cloud 2001. A moment-of-momentum flux mass air flow measurement device. Measurment Science and Technology, 12(2), N9-N13.)

This work was supported in part by the Consortium for Ultra High Efficiency Quiet Fans


Design Of A Sliding Mechanism In The Application Of Vibration Suppression Of A Nonlinear Beam

Authors: Tingli Cai; Ranjan Mukherjee; Alejandro R. Diaz

Abstract: This work presents the experimental realization of a slider to be applied in a vibration suppression system for a nonlinear beam. The slider constrains the transverse displacement of the beam locally but not the beam's rotation. Friction is designed to be minimum while the slider moves on the surface of the beam. The device measures the reaction force from the beam and prescribes sliding motion through a belt-drive actuator to do negative work on the beam. Filters are used to meet the bandwidth requirement of the actuator. The goal of this design is to stabilize the vibrating beam which is subject to various disturbance.


Multiscale Free Edge Investigation Of Composite Laminates

Authors: Christopher Cater; Xinran Xiao

Abstract: Stress gradients are present at the free edges of laminated composites as a result of property mismatch between lamina of varying orientations. This phenomenon, known as the free edge effect, has been studied extensively at the meso-scale. Typically, each lamina is assumed to be transversely isotropic and homogenous. This approach, however, neglects the heterogeneous nature of the composite at the microscopic scale and its influence on damage initiation. In this work, a multiscale approach is presented to explore the microscopic stresses, at the scale of fiber and matrix, present near the free edge. To accomplish this task, a semi-concurrent multiscale approach is implemented into ABAQUS. Two length scales are utilized: 1) the meso-scale, utilizing a composite laminate Representative Volume Element (RVE), to capture the heterogeneity of the individual lamina; and 2) the micro-scale, containing a 3D RVE of the fiber and matrix constituents. The standard kinematic relations of a computational homogenization approach are modified to allow for the use of free-edge boundary conditions at both the meso-scale and micro-scale, allowing for the investigation of free-edge effects. To simplify the problem, constituent materials are assumed linear elastic, allowing for a one-way coupling between the two length scales, similar to a global-local approach. The research addresses the validity of the solutions to the free edge boundary value problem (BVP), determining appropriate RVE sizes to achieve accurate free-edge stresses near the regions of interest. It will also present the microscopic stresses, which are potential sources of fiber/matrix debonding or matrix cracking.

This work was supported in part by Cooperative Agreement No. W56HZV-07-2-0001 between U.S. Army TACOM LCMC and Michigan State University.


Aeroacoustic Measurements Of The Self-Noise Of A Rotating Controlled Diffusion Blade

Authors: Behdad Davoudi; John Foss

Abstract: The self-noise of a 9 blade fan has been examined using an array of microphones. The array allows the principle of beam forming to be used such that extraneous noise effects are minimized. The blades are shaped as controlled diffusion airfoils with a chord length of 133.9 mm and a span of 126 mm. The nominal tip clearance is 4 mm. The pressure rise and flow rate measurements are based upon the techniques described in Morris and Foss (2001) and their representation of the Axial Fan R & D facility at MSU.

Acoustic measurements were obtained with arrays of Panasonic microphones placed above the fan plane (in the upstream flow). Microphones were located in two different circular patterns (r= 303 mm) at two different height levels, which are both directly above the blades’ mid-span. Each Panasonic microphone is calibrated with a Larson Davis microphone which is the reference microphone for the acoustic measurements. The beam forming method was employed to process the data obtained from two distinct circular arrays. Since the microphones in each array were at different distances with respect to the fan (sound source), combining the microphone voltage signals while accounting for the “source-to-microphone” distances enhances the extraction of the self-noise characteristics from the microphone measurements.

By manipulating specific operating conditions, selected based on the performance curves for each blade configuration, the relative incidence angle seen by the blades is altered, and its effect on the propagated broad band noise has been recorded.

This work was supported in part by UHEQ Consortium


Analytical Solution For Fluttering Motion Of A Free Falling Rigid Flat Plate In A Stationary Viscous Fluid

Authors: Behdad Davoudi; Indrek Wichman

Abstract: The problem of freely falling thin flat plates in a stationary viscous fluid has been previously examined. This phenomenon, for falling cards or flying leaves, is often called fluttering motion. Fluttering motion depends on specific conditions, i.e. plate/fluid relative dimensions and densities, and it can undergo a transition to tumbling motion (not discussed here). Previous studies have dealt with this problem using computational, analytical and experimental approaches. In most of the computational approaches, researchers have tried to either directly solve the Navier-Stokes equations or to utilize the circulation around the plate (for example) in various semi-empirical models. The analytical works employ a variety of force coefficients or a relation for the circulation. We follow a similar approach. Only a few studies involve experiment.

In this work, a simple analytical approach for fluttering motion is presented assuming 2-D motion (i.e., planar descent). The aerodynamic force balance is written, and a variable angle of attack based on the plate transitional velocity and direction is defined. The force balance utilizes relations for lift and drag coefficients suggested by flat plat theory. The moment coefficient around the center of the plate is obtained using an experimental correlation formula for the moment coefficient about the quarter chord. A rotational drag term is developed to take into account the drag force exerted to the plate by its rotation. Since the Reynolds number based on the plate chord is of the order of 1000 the Blasius solution is used in order to approximate the frictional drag over the plate. The remaining forces considered here are exerted by the actions of buoyancy and gravity.

The advantage of this model is that equations of motion are universal, and therefore valid throughout the fluttering motion. The numerical solutions are in reasonable agreement with the previous experiments with respect to both order of magnitude of various derived quantities and also the plate trajectory (motion) through the fluid.


Therapeutic Operation Decision For The Abdominal Aortic Aneurysm Using The Growth Measurement

Authors: H. Gharahi; B. A. Zambrano; C. Lim; J. Choi; W. Lee; S. Baek

Abstract: The maximum diameter and their growth rates are usually regarded as key factors for making a decision on therapeutic operation time for an abdominal aortic aneurysm (AAA) patient. There is, however, debate on what the better standard method is to measure the diameter and there are still remaining shortcomings among different methods. Currently, two dominant methods for measuring the maximum diameter are one on the planes perpendicular to the centerline (orthogonal diameter) and the other on the axial planes (axial diameter). Alternatively, a new robust method, named ‘inscribed sphere diameter’, proposes to measure the diameter and, additionally, addresses the similarities and discrepancies of these three methods. We investigate if the spherical diameter is a better representative of AAA volume and if it serves as a better measurement for aneurysm size in analogy with other measurements. A set of longitudinal CT scan images from nine Korean patients are investigated and compared with a previous study conducted with patients from Europe. Growth rate parameters are calculated in different diameters and the total volume and the correlations between them are studied. Furthermore, an exponential growth pattern is sought for the maximum diameters over time to examine a nonlinear growth pattern in AAA expansion, globally and locally. The results indicate that the inscribed-spherical diameter shows the most reliable and robust measurement of the diameter compared to transverse, in-plane diameters. Based on growth parameters calculated in the results, we suggest that growth parameters in the Korean and European patients are not significantly different.

Keywords: Korean patients, exponential growth, transverse diameter

This work was supported in part by National Institute of Health (NIH)


A Computational Investigation Of An Asymmetrically Pitching Airfoil

Authors: Patrick Hammer; Ahmed Naguib; Manoochehr Koochesfahani

Abstract: The recent heightened interest in the aerodynamics of flapping-wing flight is motivated by the potential for engineering unmanned micro aerial vehicles (UMAV) with unique capabilities similar to natural fliers, e.g. effective generation of lift at low Reynolds number, high maneuverability, and tolerance to wind gusts. Such UMAVs could have a wide range of applications in search and rescue operations, monitoring hazardous emissions, and reconnaissance. In this work, computations are performed for studying the flow field and accompanying forces induced by asymmetrically pitching a NACA 0012 airfoil at low Reynolds number. The majority of past research addresses sinusoidal motions, with almost none considering the influence of motion asymmetry: a parameter that is likely important for designing UMAVs. The current implementation of the flow solver utilizes a multi-grid approach to adequately resolve the flow details in the wake of the airfoil. Results are presented for two motion trajectories: equal pitch-up and pitch-down (i.e. symmetric motion) and a pitch-up that is faster than pitch-down. The computed lift and drag forces on the sinusoidally pitching airfoil show convergence for freestream Mach number below 0.025. In addition, the vortex street in the wake of the asymmetrically pitching airfoil shows two modes for vortex shedding during the slow phase of the motion depending on the Mach number: 1) two vortices that eventually merge together; 2) two vortices whose separation increases as they convect downstream. The impact of the asymmetry on the flow field and the forces is illustrated for the low Mach number of 0.025.

This work was supported in part by AFSOR Grant No. FA9550-10-1-0342.


Premixed Tulip Shaped Flames In A Rectangular Combustion Chamber

Authors: Ashwin Hariharan; Indrek Wichman

Abstract: An experimental and numerical investigation is performed of premixed flame propagation in a constant volume rectangular channel with an aspect ratio of six (6) that serves as a combustion chamber. The ignition event is followed by an accelerating convex shaped flame-front. A deceleration of the flame is followed by the formation of a concave “tulip” shaped flame-front. Eventually, the flame is extinguished because of collision with the cold wall on the opposite end of the confined channel. Numerical calculations of the combustion event are performed to understand the influence of pressure waves, instabilities, and flow field effects causing changes to flame structure and morphology. The transient 2-D numerical simulation results are compared with transient 3-D experimental results.

This work was supported in part by Advance Research Projects Agency - Energy (ARPA-E)


The Effects Of Chlorinated Compounds On Instantaneous Water Heaters

Authors: Andrew Koch; Alex Schuen; Elisa Toulson

Abstract: Bradford White Corporation requested the investigation of the impact of water with a high concentration of chlorine compounds in an instantaneous water heater. Research was done to identify chlorinated compounds found in water supplies in the United States. An analysis of the chlorinated compounds was completed to determine which chemicals are known to corrode metals, primarily stainless steel. An apparatus was built to test the instantaneous water heater with water containing the chlorinated compounds. This apparatus was designed to simulate realistic conditions for water heaters. A GCMS analysis was developed to test the degradation rate of the chlorinated compounds. Test samples with distilled water and the correct concentration of each chemical were made and heated to simulate the conditions in the water heater. The GCMS analysis results indicate minimal degradation of the chemicals in water.

This work was supported in part by Bradford White Corporation


Unit Cell Modeling To Predict Permeability For Composite Manufacturing

Authors: Timothy Luchini; Stephen Sommerlot; Alfred Loos

Abstract: Detailed geometric models can be used to accurately design molds, to predict imperfections, and correct potential design issues before more expensive resources are spent in fabrication of composites by liquid molding processes. Fiber preforms are frequently modeled by using a unit cell of repeating geometry, which allows the model to be developed in a computationally efficient manner. Using models to numerically predict permeability can reduce the need for experimentation and allow for fast characterization of large numbers of fiber preforms. The research described here validates this by using the unit cell approach to model in-plane saturated permeability and comparing the predictions with experimental results obtained using a custom test fixture to measure permeability of a plain weave S glass fabric. The fabrics are characterized using a scanning electron microscope (SEM) and the finished composites are studied at the desired volume fraction for material characteristics. The actual fabric geometry is important in the geometrical prediction of permeability, and the model is generated based on a set of python scripts defined in an open source textile software package called TexGen. A unit cell computational fluid dynamics (CFD) model for permeability prediction is presented, which considers a dual scale preform that uses Stokes flow in the inter tow voids and Gebart’s permeability prediction in the fiber preform. The CFD results produced with the commercial package Fluent are shown to correlate well with experimental results, and this research further validates the results of the open source software with data produced by commercial fluids solvers.

This work was supported in part by GE Aviation


Crystal Plasticity Modeling Of Deformation Of Ferrite And Martensite Micropillars In A Dual Phase Advanced High Strength Steel

Authors: Aboozar Mapar; Taejoon Park; Farhang Pourboghrat

Abstract: Advanced high strength steels (AHSS) have been widely used in auto industry over the past few decades. These steels, without compromising their mechanical properties, have made a considerable weight reduction in cars. The weight reduction, increases the fuel efficiency of the automobiles and makes them more appealing to customers. Many of AHSS contain multiple phases. One needs to know the behavior of each phase in order to understand the deformation behavior of the multiphase steel.

Ferrite and Martensite phases respectively have BCC and BCT crystal structures. The classical crystal plasticity model; however, was initially developed for FCC material, which has close packed planes and 12 distinct slip systems. The active slip systems in a FCC crystal can be found from the Schmid law, which states that dislocation slip occurs when the shear stress on a slip system parallel to the slip direction reaches a critical value. This law is not valid in BCC and BCT materials, which do not have close packed planes. In these materials, stress on non-planar or non-parallel to the slip direction can affect the initiation of dislocation slip. This is known as non-Schmid behavior.

In this study, a non-Schmid crystal plasticity model was developed and implemented into a commercial FEM software (Abaqus) as a user defined subroutine. This model was then used to simulate the compression behavior of Ferrite and Martensite micropillars. The results of the experiments and simulations will be discussed in the poster.


Control Of Hybrid Dynamics With Application To A Hopping Robot

Authors: Frank Mathis; Ranjan Mukherjee

Abstract: Control of dynamic motion is a crucial area of study in robotics as on frequently wants the robot to behave in a desired motion pattern rather then moving to a set point. Furthermore, the motion of the robot commonly involves changing dynamic behaviours commonly due to environmental effects such as surface contacts, which leads to hybrid dynamic systems. A common area of such hybrid dynamic control is in legged robots which have hybrid dynamic behaviour such as switched dynamics due to changing legs and impulsive dynamics due to ground contacts, but also require control to a dynamic trajectory defining the walking or running motion. For this, The spring loaded inverted pendulum (SLIP) model is commonly used as a simplified model to describe the dynamic motion. Based on this model, the control of hopping robots has been widely investigated. A fundamental limitation of the model is that it fails to account for impact with the ground, and this is due to its single degree-of-freedom in the vertical direction. Here we investigate the dynamic control of a four-link hopping robot. The advantage of our method is that the entire dynamics of each robot is considered for the control design allowing the controller to compensate for the impulsive dynamics as well as higher order behaviour which are unaccounted for in simplified models such as the SLIP model.


Pore Formation And Deformation In Membrane Bilayers

Authors: Vahid Mirjalili; Nikolai Priezjev; Michael Feig

Abstract: In this study, we have developed a new method called Density Biasing method, that under molecular dynamics (MD) framework can be used to effectively form pore and one-sided deformation in membrane bilayers. The density biasing method tries to increase the density of water molecules in a cylinder aligned to bilayer normal axis. Using this method, we evaluated the free energy cost of forming a pore in the bilayer, and how the presence of hydrophilic compounds in bilayer center affect this process. It is found that under the presence of acetamide in bilayer center, two equi-stable states (flat bilayer and deformed state) exist, which are separated by a free energy barrier. The flat bilayer state is stabilized by penetration of a single water molecule.


Analyses Of Hydrodynamic Features And Separation Performance Of A Hydrocyclone Used For Oil-Water Separation

Authors: Abdul Motin; Volodymyr V. Tarabara; Charles A. Petty; André Bénard

Abstract: This research addresses critical aspects of the hydrodynamics and separation performance of a de-oiling hydrocyclone used in produced water treatment, oil spills cleanup, as well as refining of petroleum products. Fundamental hydrodynamic features that influence separation performance of a de-oiling hydrocyclone are examined. Velocity and pressure inside the hydrocyclone are calculated by numerically solving Reynolds Average Navier-Stokes equations closed with Reynolds Stress Model. Separation efficiency of the hydrocyclone is estimated by analyzing trajectories of dispersed oil droplets in the flow where the trajectory is calculated by solving the equation of motion and the force balance on a dispersed droplet. Effects of hydrocyclone geometry and the inlet conditions on the internal flow structure, short circuit flows, vortex core pattern, and trajectories of droplets are investigated based on computational fluid dynamics. Results indicate that the conventional de-oiling hydrocyclones have a finite turndown ratio i.e. it exhibits acceptable separation efficiency only for a certain range of the Reynolds number. Internal flow structures provide a fundamental understanding for possible approaches to redesign a hydrocyclone.

This work was supported in part by U. S. Environmental Protection Agency (award no. RC 83518301); U.S. National Science Foundation (award no. IIA-1243433); Michigan State University Foundation (Strategic Partnership Grant 71-1624).


Surface Pressure Measurements From Multiline Single-Component Molecular Tagging Velocimetry

Authors: David Olson; Ahmed Naguib; Manoochehr Koochesfahani

Abstract: Knowledge of the aerodynamic forces acting on an object in relative motion to a fluid is of paramount importance for the structural design of the object, for determining the payload, if the object is of the lifting type, and for minimizing the energy consumption in moving the object relative to the fluid. Traditional experimental methods of obtaining these forces require either considerable embedded instrumentation or intrusive hardware in the flow. An ideal measurement technique would provide a non-intrusive pressure and shear stress distribution around the body without the expense and limitations of fixed embedded instrumentation. This study considers the feasibility of estimating the surface pressure distribution and shear stress based on high-resolution single-component molecular tagging velocimetry. The method relies on the connection between the surface pressure gradient and the second order wall-normal derivative of the velocity component tangent to the wall. We show the application of this approach to measuring the surface pressure and shear stress distribution on a circular cylinder in cross flow at Re = 6,000. Results compare favorably with data in the literature.

This work was supported in part by AFOSR grant number FA9550-10-1-0342.


Blood Flow Responses To Loading: Interpretations For Skin Ulcers

Authors: Wu Pan; Josh P.Drost; Marc D.Basson; Tamara Reid Bush

Abstract: Skin ulcers are a significant health concern affecting the elderly, diabetics, individuals with vascular disease and amputees resulting in deep penetrating wounds. One of the factors under investigation for increased risk of skin ulcers is post occlusive reactive hyperemia injury. Such aggressive increase of blood flow in the vessel has been shown in animal studies to be related to tissue damage.

The goal of this research was to determine differences in reactive hyperemia responses across a population of healthy individuals, and individuals with chronic leg wounds.

A custom designed force applicator was adopted for different loading applications and was accommodated with Laser Doppler probe allowing loading to occur around the perfusion probe. Twenty-one patients with wounds and twenty healthy individuals participated in the tests.

The Absolute Reactive Hyperemia Value (ARHV), i.e. the maximum value of perfusion after load release, and the Relative Reactive Hyperemia Magnitude (RRHM), i.e. the difference between ARHV and the perfusion value during the loading, were compared across all three categories. Results showed that wounded legs have the highest ARHV and RRHM values followed by non-wounded legs of patients with wounds, and healthy legs had the lowest values. Statistically significant differences occurred between all groups for both normal and combined loading

Understanding changes in blood flow responses across healthy patients and patients with skin ulcers is critical to improving our understanding of skin wound formation, developing better preventive measures, and providing inputs for tissue injury models.


Equilibria Analysis And Wave Propagation In Nonlinear Chains

Authors: Smruti Panigrahi; Brian Feeny; Alejandro Diaz

Abstract: We present the dynamics of nonlinear structures with both quadratic and cubic nonlinearities. We first studied the bifurcations of equilibria of a two-degree-of-freedom twinkling oscillator in order to investigate the capacity to harvest energy. Twinkling occurs when the nonlinear structure is loaded slowly and the masses snap through, converting the low frequency input to high frequency oscillations. The kinetic energy born in these oscillations can then be harvested. We also studied the propagation of waves in an infinite nonlinear chain to highlight the effects of quadratic nonlinearity. When multiple waves pass through the chain we observe their interactions and exchange of energy in the sub and super-harmonic resonance conditions. Studying the dynamic behaviors of these structures is an important step toward understanding the capacity for energy harvesting, energy dissipation and crashworthiness, vibration isolation, vibration absorption, event detection, and nonlinear waveguides.

This work was supported in part by National Science Foundation


Development Of Soft Nanoimprint Lithography And Its Application In Nanowire Fabrication

Authors: Snehan Peshin; Junghoon Yeom

Abstract: As an unconventional lithographic technique with high throughput patterning at great precision and low cost, soft nanoimprint lithography has a great scope for fabricating nanowires and other nanostructures in combination with metal assisted chemical etching. Traditional lithographic approaches use photons or electrons thus incorporating the limitation of wavelength and beam scattering. But in soft nanoimprint lithography, there is a direct mechanical deformation of a polymeric film on the substrate by impressing with soft elastomer replica typically made of PDMS. We are developing a process to replicate micro- and nanostructures from a soft master mold by imprinting on a UV-curable polymer-coated substrate. This process is facile, reliable, and scalable hence inexpensively creating an array of patterns of various geometries. For example, an array of dot patterns has been fabricated by soft nanoimprint lithography, and metallization followed by the lift-off process produces a metal film with ordered holes. Now a noble metal film such as Au or Ag can catalyze the etching reaction with the silicon substrate in the presence of oxidant (e.g. H2O2) and HF. This combined technique creates a new opportunity to cheaply fabricate nanowires. These nanowires will find applications in gas sensors and various optoelectronic applications. The focus of the presentation will be placed on challenges related to producing good quality imprints and various parameters for optimized quality as well as time factor.


Generative Variable Length Genetic Algorithms

Authors: Matthew Ryerkerk; Ron Averill; Erik Goodman; Kalyanmoy Deb

Abstract: Optimization algorithms typically operate with a fixed-sized genome. However, there exists a class of problems where the number of design variables may not be fixed. Such problems include sensor placement, laminate composite design, packing, or neural network problems where each solution can be defined with a range of possible component numbers. Gradient based algorithms are ill-suited for such problems. Genetic algorithms are viable candidates, however the traditional operators are of little use with a variable-size genome. Our previous work has resulted in the development of a variable-length genetic algorithm (VLGA), capable of determining solutions of the optimal size without a priori knowledge. VLGA was shown to be very effective for several testbed problems, however it requires a high number of evaluations, which limits its usefulness for real world engineering problems. The variable-size problems that we have studied are all highly multimodal, many solutions have similar fitness but dissimilar topologies. The current algorithm, and associated operators, relies on a direct encoding of the solution into the genome. This makes it near impossible to effectively explore different topologies in the design space, to move from to another would require coordinated changes to many components at once. Instead a generative, indirect encoding, and appropriate operators, for variable-size problems is proposed. Small changes to this encoding could produce a coordinated change to all components in a solution, facilitating a fast and effective search over many topologies.

This work was supported in part by BEACON - Center for the Study of Evolution in Action


Prior Distributions Of Material Parameters For A G&R Computational Model Of Abdominal Aortic Aneurysm

Authors: Sajjad Seyedsalehi; Liangliang Zhang; Jongeun Choi; Seungik Baek

Abstract: Advances in computational modeling of the aorta in line with the use of subject-specific data have promised a growing potential in clinical diagnosis and treatments of vascular diseases. In enhancing an accurate prediction of the vascular disease progression, there is, however, an important need for a systematic tool towards the patient specific modeling. Particularly, considering the intra-patient variation of model parameters, the prior distribution has strong influence on computational results for arterial mechanics and one crucial step towards patient specific modeling is the use of patient specific model parameters instead of used population averaged values. To this goal, we present exploiting a new statistic tool of an arterial model, as a better systematic tool, and then estimate the prior distribution for the model parameters, using experimental results for 17 healthy abdominal aortas. We investigate the correlation between estimated parameters with noninvasively assessable parameters, age and gender. Using the correlation results with the application of Box-Cox transformation, we construct the conditional joint distribution and confidence intervals for model parameters. This information improves the prior distribution of subject-specific model through specifying parameters using age and gender. A Bayesian based statistic tool decreases the uncertainty and error in the prediction of subject-specific model as presenting aortic material behavior. The results from this study will be used as the prior information necessary for the statistic tool of G&R.

This work was supported in part by National Science Foundation (NSF)


An Interfacial Debonding-Induced Damage Model For Graphite Nanoplatelet Polymer Nanocomposite

Authors: Azadeh Sheidaei; Farhang Pourboghat

Abstract: In situ tensile tests show damage initiates in nanocomposite mainly by interfacial debonding. In this paper a hierarchical multiscale model is developed to study the damage initiation in polymer /graphite nanoplatelet (GNP) composites. The cohesive zone model has been adopted to capture the nanofiller deboning. The results of atomic simulations of GNP pullout and debonding tests have been used to obtain the traction-displacement relation for cohesive zone model (CZM). The effects of volume fraction and aspect ratio of the GNP and strength of the interfacial adhesion on overall stress-strain response of the nanocomposite have been investigated. Results show debonding have a significant effect on overall stress-strain response when volume fraction and aspect ratio increase. The results also indicate that GNP/polymer interfacial strength plays a key role in damage mechanism of nanocomposite.


A New Continuum Damage Mechanics Model For Crash Simulation Of Fiber Reinforced Composites

Authors: Danghe Shi; Xinran Xiao

Abstract: Fiber reinforced composites are widely used in aerospace, automotive industry due to their high stiffness, strength-to-weight ratio, corrosion resistant and energy absorption ability. However, their complex failure mechanisms make it very difficult to analytically and numerically model their behavior under crash and thus limited their application for massive productions. Many works have been attempted to simulate the crashworthiness of composite structures, particularly to evaluate the deformation behavior and to determine the energy absorbing efficiency of various composite structures. However, the existing simulation models generally need to introduce many non-measurable parameters which limited their practical applications. In order to solve this problem, this work focused on the implementation and development of a thermodynamically consistent continuum damage mechanics (CDM) model. All the parameters needed in this model can be determined by experiment. It was proved that this model is able to capture the behavior of several different fabric forms of fiber reinforced composites during crash including some special event like hysteresis phenomenon.


Modeling Advancing Flow Fronts In Composite Manufacturing

Authors: Stephen Sommerlot; Timothy Luchini; Alfred Loos

Abstract: Flow front propagation in liquid composite molding (LCM) is subject to the fiber architecture, resin properties, and the preform geometry being infused. Defining accurate resin flow fronts for complex fabrics is challenging as the permeating fluid often progresses in an uneven “fingering” manner due to variable porosity and flow channels the textile geometry creates. Modeling the advancing flow fronts in LCM accurately is important as air entrapment and void formation can arise due to the resin and air phase interaction at the flow front. These are important considerations for mold design, prediction of voids, and finished part evaluation. The research described here presents simulations of the advancing resin flow front through textile preforms. A tetrahedral and voxel based mesh is generated from TexGen scripts modeled from a unit cell of textile geometry and comparisons are made showing mesh independence for a sufficiently refined mesh. A computationally efficient, unit cell based, solution is shown to account for the intricacies of the air to resin phase transition seen in actual mold fills. A multiphase, transient finite volume solution is found in Fluent using the generated mesh, and post-processed to produce an advancing flow front simulation. For model validation, a plain weave S-glass fabric is first characterized geometrically for TexGen input parameters. Then, experimental advancing flow front infusions are conducted with an instrumented line-source to line-sink visualization fixture. Model and experimental flow fronts are overlaid yielding favorable qualitative results.

This work was supported in part by GE Aviation


An Energetically Enhanced Plasma Ignition System For Use In Internal Combustion Engines

Authors: Bryce Thelen; Elisa Toulson

Abstract: The effects of a plasma enhanced ignition system on the performance of a small single cylinder four-stroke gasoline engine are examined. Dynamometer testing of a 34 cc gasoline engine is performed comparing a stock ignition coil with a radio frequency plasma ignition system. The radio frequency system is designed to provide a quasi non-equilibrium plasma discharge that features a high voltage pulsar capable of providing voltages of up to 30 kV and 400 mJ of energy per discharge. Tests show improvement of the engine’s performance in regards to combustion stability and tolerance of lean air fuel mixtures with the radio frequency system. Additionally, high speed images of the radio frequency system taken in a different 0.4 liter optical engine are presented.

This work was supported in part by the Air Force of Scientific Research under Reward FA9550-10-1-0556.


Numerical Simulations Of Turbulent Jet Ignition And Combustion

Authors: Abdoul Ahad Validi; Abolfazl Irannejad; Farhad Jaberi

Abstract: The ignition and combustion of a homogeneous lean hydrogen-air mixture by a turbulent jet flow of hot combustion products injected into a colder gas mixture are studied by a high fidelity numerical model. Turbulent jet ignition can be considered as an efficient method for starting and controlling the reaction in homogeneous combustion systems used in advanced internal combustion and gas turbine engines. In this work, we study in details the physics of turbulent jet ignition in a fundamental flow configuration. The flow and combustion are modeled with the hybrid large eddy simulation/filtered mass density function (LES/FMDF) approach, in which the filtered form the compressible Navier-Stokes equations are solved with a high-order finite difference scheme for the turbulent velocity and the FMDF transport equations are solved with a Lagrangian stochastic method to obtain the scalar (temperature and species mass fractions) field. The hydrogen oxidation is described by a detailed reaction mechanism with 37 elementary reactions.


Rotary Induced Impact Testing And Analysis Of Composite Fan Cases

Authors: Andy VanderKlok; Jim Dorer; Xinran Xiao

Abstract: High speed fans play a vital role in the automotive, aeronautical, and medical fields. In the aeronautical field alone these fans are commonly used to produce thrust for commercial and private airlines. It is well understood for these airliners to sustain ground velocities near Mach I speeds, the engines themselves must turn a fan at very high rpm to gain the thrust needed to propel the aircraft to such velocities. Because of the high energies involved with such fast rotating machinery, failure known as a fan blade out event (FBO) can occur often caused from bird strike or fatigue. When this happens debris is released at ballistic rates of speed and can cause catastrophic damage to nearby aircraft components. In addition, the release of a blade creates a severe imbalance of the shaft leading to eccentric motion. This motion causes the remaining blades to rub the fan housing causing further damage. Post FBO shaft imbalances can negatively affect other engine components such as the shaft, bearings, and rotors. Properly characterizing the dynamics and failure modes is crucial in understanding what happens during this event. Although this is difficult to replicate experimentally at reasonable cost; it can be with a spin pit. Utilization of a spin pit will more closely be representative of the loading conditions present during an actual FBO. The pit allows for experimental high speed video and strain data acquisition for analysis of a controlled FBO for composite blades or fan casing combinations.

This work was supported in part by NASA


A Microstructure-Resolved Model For Li-Ion Battery With Silicon Anode

Authors: Miao Wang; Xinran Xiao

Abstract: High capacity anode materials such as silicon experience large volumetric changes during charge/discharge cycling in battery cells. Large cyclic deformation often leads to particle fracture, mechanical failure and delamination at particle-binder, particle-current collector interfaces, which results in pulverization and capacity fading. An electrode microstructure resolved full-cell model has been developed using COMSOL Multiphysics to investigate the kinetics of Li transport and electrochemical reactions, stress accumulation and structural deformation. This model is further adopted for high capacity anode material silicon by considering elastic moduli, yielding strength, partial molar volume and Poisson’s ratio variations along with lithiation/delithiation. The model validation is ongoing by comparing the computed cell voltage, stress and volume expansion with cycles with real-time experimental results. This model allows the observations of morphologies and stress distributions in silicon anode and the influence of volume variation on the electrochemical reaction of the cell. It will serve as a design tool for structured silicon lithium-ion batteries.

This work was supported in part by National Science Foundation


The Softening Behavior Of A Polymeric Battery Separator In Solutions

Authors: Shutian Yan; Yue Qi; Xinran Xiao; Xiaosong Huang

Abstract: The mechanical integrity of the separator is crucial to the performance, abuse tolerance, and durability of Li-ion batteries. It has been observed that, when tested in electrolyte solvents, the elastic modulus of a polypropylene (PP) separator Celgard 2400 reduced to about a half of the value obtained in air or water. The separator regained its modulus in air after being dried. This recoverable softening response suggests that the behavior is induced by the solvent. The models considering the change of surface tension, however, cannot explain the magnitude of the softening. This work investigated the problem using atomistic modeling. The PP separator has a porous microstructure formed by patches of crystalline phases connected by nanofibers. To capture the mechanical responses of these phases in different environments, atomistic models for crystalline and amorphous PP nanofibes were built separately, and the molecular dynamics (MD) simulations were performed in vacuum, water and a typical electrolyte solvent dimethyl carbonate (DMC). The results showed little interaction in all cases except for the case of amorphous PP in DMC. DMC molecules penetrated into the amorphous PP nanofiber, reduced the local density and the elastic modulus. The results suggest that the softening phenomenon may be attributed to the strong attraction of the electrolyte solvent molecules with the amorphous fibrous PP regions of the separator.

This work was supported in part by NSF and GM


Predictive Boundary Management Control Of A Hybrid Powertrain

Authors: Jie Yang; Guoming G. Zhu

Abstract: Hybrid Electric Vehicles (HEV) is capable of improving fuel economy with reduced emissions over traditional vehicles powered by internal combustion engine alone. However the HEV vehicle durability is significantly limited by the the useful battery life, and the battery life could be significantly reduced if it was operated at its allowed charge or discharge limits, which could occur especially at extremely low battery temperature, leading to permanent battery damage and reduced battery life. In order to extend the battery life, this paper proposed a battery boundary management control strategy based upon the predicted desired torque to proactively make the engine power available to reduce future battery over-discharging. The proposed control strategy was validated in simulations and its performance was compared with the baseline control strategy under Federal Test Procedure (FTP) and other four typical city and highway driving cycles. The simulation results show that the proposed control strategy is very effective when the battery temperature is under zero degree, and the over-discharged power are reduced around 65% under aggressive US06 and ARB02 driving cycles, 45% under highway and city FTP and NYCC city driving cycles, and 30% under , highway IM240 driving cycle, respectively.

This work was supported in part by DOE funding through Chrysler, LLC


Effect Of The Intraluminal Thrombus Layer On The Diameter – Expansion Relationship In Abdominal Aortic Aneurysm (AAA): A Longitudinal Patient Study

Authors: Byron A. Zambrano; Jongeun Choi; Seungik Baek

Abstract: Abdominal aortic aneurysm (AAA) is the permanent focal distention of the aorta and its end stage is related to death with its rupture. The rupture occurs when the wall stresses on the wall overcome by the wall strength. These stresses, according to the law of Laplace, are positive correlated to the diameter. Additionally, the diameter is positively correlated to AAA expansion rate. In other words, the larger the diameter, the larger the expansion, and the higher the stresses on the AAA wall, potentially leading to rupture. Unfortunately, these two parameters are not always reliable. This lack of reliability might be due to the presence of other factors that may contribute to AAA prognosis, like the intraluminal thrombus layer (ILT), that are not considered in the prediction. ILT buildup is found covering partially or fully the lumen wall of 75% of AAAs, increasing in prevalence as aneurysms enlarge. Hence, this study uses longitudinal CT images of 9 different patients to study the effect of the ILT in the diameter-AAA expansion relationship. Our results showed that when ILT is present, the strength of the expansion-diameter relation is not only weakened, but the nature of this relation is also altered since expansion is accelerated compared to patients without ILT. This suggests that a distinction between patients with and without ILT should be made before predicting rupture or programing the follow up time interval.

This work was supported in part by National Institute of Health (1R21HL113857-01).