Abstracts Submitted from Mechanical Engineering
Undergraduate Summer Research Symposium August 12, 2009

Ordered alphabetically by student's last name

Adelmann Carrera
Hennigan Liang Pretz
Aten Che Hilgart Macasevich Russakow
Baker Friedrich Kee Melrose Russo
Bonnevie Gallo Kistler Mulrooney, Th Schoepflin

Installation and  Configuration of an Experimental Testbed for Robot Navigation in Indoor Environments

Joseph Adelmann, Nathaniel Dylan Kee, and Herbert Tanner
Department of Mechanical Engineering

The future of robotics does not lie in controlling a single robot but in effectively controlling large groups of robots. This can be attained through computer programs. However before using these programs on actual robots, to save time and energy it is more practical to simulate the results. Player and Gazebo, of the open source robot testing software Player Project, are used to accomplish this feat. Player allows users to write robot control programs in any programming language. Gazebo is used to simulate multiple robots in a three dimensional world utilizing rigid-body physics.  After simulation in Player and Gazebo, the actual code is implemented wirelessly on CoroBot. CoroBot is a robot manufactured by CoroWare, Inc. specifically for research purposes. Due to inaccuracies in the simulation with respect to real world such as wheel slippage, CoroBot's movements do not mirror the result of the simulations. Therefore,  a Vicon MX motion capture system consisting of eight infra-red cameras is utilized to update CoroBot on its location. This system tracks the motions of the robots relative to an established origin. The bulk of the work was spent installing the necessary software and hardware. Then all the systems had to be interfaced with one another wirelessly. Initial control programs focused on the robots interacting with their environment i.e. recognizing obstacles. Future endeavors will utilize this testbed to simulate and control larger, more diverse groups of robots including both ground and aerial robots.

Tensile testing of Polymer Electrolyte Membranes in varying environmental conditions

Alex B. Aten, Michael H. Santare, Anette M. Karlsson,Tom Cender, Melissa Lugo, and Andrew Fassler
Department of Mechanical Engineering

Polymer Electrolyte Membranes (PEMs) perform an essential function in the chemical reaction in a fuel cell.  PEMs conduct protons, but force electrons to travel through a circuit in order to complete a chemical reaction.  This creates an electric current that can be used as power.   As a fuel cell is used, the membranes wear down and eventually fail.  Mechanical degradation is an important contributor to the failure of PEMs in fuel cells.  In an operational fuel cell, the PEM is exposed to a wide range of temperatures and humidity.  The membranes swell and contract with changes in temperature or humidity.  The mechanical properties also change depending on these conditions.  The stresses that develop as the membrane is exposed to changing conditions lead to decreased durability and eventually failure.  In order to design longer lasting fuel cells, it is necessary to understand the mechanical properties of the PEMs in a variety of conditions.  This research is focused on examining the behavior of the membrane in cold (5 to -40 degrees Celsius) conditions.  By conducting a simple tensile test across a range of temperatures and humidities we can see how the Young’s Modulus and Proportional Limit Stress change with conditions.

A high-performance cathode for solid oxide fuel cells
Andrew M. Baker, James L. White, and Joshua L. Hertz
Department of Mechanical Engineering

In solid oxide fuel cells (SOFCs), ceramics are used as components because of their ability to conduct electrons and/or ions. Cathodes in SOFCs are used to facilitate the reduction of oxygen at the interface in the fuel cell. Currently, cathodes are a composite made from two materials: one which conducts electrons and one which conducts oxygen ions. These materials can be replaced with a single material known as a mixed ionic electron conductor (MIEC), which is capable of simultaneously conducting both electrons and ions. Though conventional powder processing can be used to create a ceramic MIEC, significantly more efficient MIEC’s can be formed using nanoporous ceramic processing.  A nanoporous thin film will be created by sputtering, a physical vapor deposition process. Processing and simultaneously sputtering a target of a known ceramic MIEC, barium strontium cobalt ferrite (BSCF), with a target of a polymer, polytetrafluoroethylene (PTFE), and subsequently decomposing the polymer will create a nanoporous thin film of BSCF. The increased surface area of the film, due to the porosity, will increase available reaction sites of the material and improve its ability to efficiently exchange oxygen with the atmosphere. These processes can be further optimized to create a highly efficient material which can be used as a cathode in a SOFC.  In this poster, we will describe our initial efforts to create this nanoporous film.

Microtribological Characterization of Articular Cartilage

Edward D. Bonnevie, Pierre Yao Koffi, Liyun Wang, and David L. Burris
Department of Mechanical Engineering

Osteoarthritis (OA), a degenerative joint disease associated with the degradation of articular cartilage, is the leading cause of chronic disability in the United States.  OA is difficult to detect (especially in the early treatable stages) and treat. Currently, histology techniques are used to study damage and damage mechanisms; while these destructive techniques are sensitive to local areas of subtle damage, they preclude continued studies of the specimen.  In-vivo MRI holds promise for studies of damage progression, but its resolution limits use to cases of severe and irreparable tissue damage. Joint lubrication is known to be highly sensitive to the unique biphasic structure of the cartilage matrix and its fluid pressurization response to loading.  It is hypothesized that structural damage interferes with lubrication and increases friction.  In this study, we explore the use of tribological measurements for non-destructive diagnosis of cartilage damage and spatial tracking of damage progression.  The design of this instrument is discussed in the context of the unique lubrication mechanisms of cartilage.  Contrary to suggestions in the literature, preliminary control testing with healthy cartilage has demonstrated sustained physiological lubrication during microscale contact conditions ( < 0.1).  Additionally, the friction coefficient was found to be sensitive to speed, load, path-length and probe radius.  These preliminary findings strongly suggest that i) initial localized surface damage will have an effect on the friction coefficient; and ii) microscale tribometry will allow for early detection, spatial tracking of damage progression and mechanically induced damaged. 

Optofluidic Microscopy Employing In-Line Holography

Allan Che, Caitlin Pretz, Takashi Buma, and Kenneth Barner
Department of Mechanical Engineering

The modern medical field relies on high-resolution biological imaging systems, however most of these systems still involve bulky and expensive equipment or several days of waiting for results.  The need for more compact and less costly tools without the sacrifice of image resolution is growing.  In this poster presentation, we discuss imaging systems created from inexpensive items (laser, LED, webcam, etc.) rather than from comparatively costly parts.  With the help of holographic image reconstruction, our imaging systems were able to reveal detail as small as 10 um while maintaining a small instrument size.

Influence of Processing on the Electrical Properties of Carbon Nanotube/Epoxy Composites

Sarah Friedrich, Limin Gao and Erik Thostenson
Department of Mechanical Engineering

Carbon nanotubes have drawn increasing attention for their possible uses in composite materials engineering. Their small size brings reinforcement structures down from the micrometers to the nanometers, and their exceptional physical, electrical, and thermal properties offer exciting possibilities for composite materials. In particular, the conductive properties and large aspect ratio (length/diameter) of carbon nanotubes allow the formation of conductive networks in epoxy polymer materials at concentrations below 0.1 weight percent. The calendering approach to CNT dispersion offers a scalable and cost-effective method to maximize dispersion while minimizing damage to the nanotubes. This research studies the formation of electrically conductive networks during the calendering process. Microscopy reveals that calendering the mixture at subsequently smaller gap settings is most highly effective at breaking up the nanotube agglomerates, but the electrical data showed a more complex relationship between calendaring and conductivity. Preliminary results suggest that some larger agglomerates mixed in with the highly dispersed nanotubes may increase conductivity by acting as nodes off which other conducting paths may branch.   This work is funded by the US Air Force Office of Scientific Research (FA9550-06-1-0489), Dr. Byung-Lip Lee, Program Director.


Gerard J. Gallo, Limin Gao, and Erik T. Thostenson
Department of Mechanical Engineering Center for Composite Materials

Increasing applications of fiber-reinforced composites into large-scale structures requires the need to develop reliable damage sensing techniques. It has been found due to the conductive nature of carbon nanotubes that infusing them into the matrix of a fiber composite creates a sensory network from which the electrical resistance of the specimen can be acquired. Increases in the electrical resistance indicate the development of damage in the composite. For testing, multi-walled carbon nanotubes were dispersed into an epoxy resin and infused into both non-conductive glass-fiber preforms as well as highly conductive carbon-fiber preforms. Both sets of specimens underwent cyclic loading tests during which stress, strain, and changes in electrical resistance were recorded. A crack density parameter was developed for the glass fiber specimens using edge replication. The increase in electrical resistance during each loading phase of the test cycles was compared with the corresponding crack density in order to display the various damage stages before failure. Electrical resistance data collected from the carbon fiber specimens was evaluated in order to isolate the resistance data of the carbon nanotubes from that of the carbon fiber. This separation would provide a better understanding of the location of damage in the composite.  This research is sponsored by the Air Force Office of Scientific Research (Dr. Joycelyn Harrison, Program Director) and Acellent Technologies Inc. under Subaward 090000288

Performance of Carbon Foam as a Gas Diffusion Layer in a PEM Fuel Cell

Bryan E. Hennigan, Feng-Yuan Zhang, Ajay K.  Prasad, and Suresh G.Advani
Center for Fuel Cell Research  and Department of Mechanical Engineering

Proton exchange membrane fuel cells (PEMFCs) have great potential as energy sources for portable, automobile, and stationary applications. High initial cost and degradation overtime have been among the biggest obstacles for PEMFC commercialization. One possible area for savings is in the material selection of gas diffusion layers (GDLs), which perform important functions in fuel cells, such as distributing reactant gases, transporting electrons, and removing water and heat.  Carbon foam is a promising material for application as a GDL, since it possesses a unique combination of low fluid resistance, large surface area, high electrical conductivity, and relatively low cost.  Testing the performance of the highly porous, reticulated vitreous carbon (RVC) foam will demonstrate the applicability of this inexpensive material as a GDL. Carbon foams with a range of porosities and densities will be characterized and tested under a number of different operation conditions, and the results will be compared with conventional GDL materials such as carbon paper.

Using Robots for Self-Generated Mobility in Infants

Dave Hilgart, Zachary Schoepflin, James Galloway, Ji-Chul Ryu, and Sunil K. Agrawal
Department of Mechanical Engineering

Mobility is a necessary aspect of human development.  Humans require the ability to move about in order to function and interact with their surroundings and with other individuals.  Mobility begins to develop in the early stages of life, when toddlers begin to crawl and eventually walk.  Certain disorders and disabilities, e.g. cerebral palsy, spina bifida, and Down syndrome, can cause children to develop atypically, hindering their motor development.  This impaired physical development can affect social and cognitive development as well.  This project aimed to provide atypically-developing, young infants with self-generated mobility using engineering aspects of mobile robotics.  Results suggest that both typically- and atypically-developing infants will move independently using a mobile robot, even without any formal training.  Funding provided by the University of Delaware Undergraduate Research Program and the National Science Foundation.

Substituting Recycled Cooking Oil in an IC Engine:

 Feasibility Study and Thermal Control Implementation

William “Jay” Kistler, Doug Brunner, Suresh Advani, and Ajay K. Prasad
Center for Fuel Cell Research / Department of Mechanical Engineering

The goal of this project was to study the feasibility of substituting used cooking oil as a fuel for an internal combustion diesel engine. The primary concern is that cooking oil has a higher viscosity than diesel fuel. An experiment was first conducted to find the temperature at which the viscosity of the oil matched that of diesel fuel.  Tests with used oil samples from UD dining services and several local restaurants showed that this temperature range was 230°C to 250ºC.  The temperature was capped at around 250°C to prevent polymerization of the oil as it enters the engine.  Several chemical methods of removing the free fatty acids were attempted, but it was concluded that the quality of the samples was already adequately high.   After confirming that used cooking oil could substitute for diesel after particles were filtered out and the proper temperature was maintained, the next task was to design a heat exchanger to heat the oil using the engine exhaust gas.  Using heat transfer principles and computer simulations an optimum heat exchanger was designed.  The final design extracts heat from the exhaust and transfers it into a piece of 40 pores-per-inch Aluminum metal foam to heat the oil as it flows through.  The installation of this heat exchanger will enable the IC engine to run on used cooking oil, thereby displacing diesel and making effective use of used cooking oil.

Robotic Assistive Crawling Device for Infants with Disabilities

Sherry Liang, Stephen Dolph, Chen Xi, and Sunil K. Agrawal
Department of Mechanical Engineering

Many children with special needs, such as Down Syndrome, cerebral palsy, and autism, experience self- generated mobility problems- such as weak musculature and poor coordination. Consequently, parents and guardians tend to discourage crawling by keeping children in chairs or standing frames, primarily for safety concerns. These children spend less amounts of time to no time on the floor compared to typically developing children their age. This contributes to delays in their cognitive, perceptual, social, and emotional development that is directly associated with locomotion. To allow disabled children to reach childhood milestones concurrently with their peers, a robotic device equipped with distance sensors was developed for safe exploration. Infants can maneuver the robot in a prone position through different driving methods that utilize computer USB wireless optical mice. These devices will be easier to handle for infants as young as six months as opposed to the conventional joysticks found in power wheelchairs. Further studies will determine whether infants can be trained to drive these robots safely and what the effects of self-generated locomotion are on their long-term motor and cognitive development.

Bio-Based Composite Bridges

Kyle T. Macasevich, Jonathan O. Carrera, and Harry W. Shenton III
Department of Civil and Environmental Engineering

Composites, such as fiber reinforced plastics have taken hold of the bridge building industry as an alternative, lightweight construction material.  Currently bridges are made of concrete, steel, and wood which can be expensive and are less environmentally friendly.  The development of a composite material from renewable resources would tackle the issue of environmental sustainability experienced with both conventional building materials and new-age composites.  This research is an analysis of the viability of bio-based composites for bridge building applications.  Tests were conducted on different natural fibers which allowed for the determination of the best fiber reinforcement for use in the bio- composite.  Sisal twine had the highest average ultimate strength, as compared to jute twine and burlap, at 18.3 ksi.  The vacuum assisted resin transfer molding (VARTM) process was used to fabricate the bio composites, made with Accrylated Epoxidized Soybean Oil (AESO) resin, recycled paper, and sisal twine.  A 20 in x 1.75 in x 2.25 in preform beam was constructed of 15 layers of recycled paper and two layers of ten sisal strings (avg diameter 0.05 in) on top and bottom.  Under three-point bending over an 18 in span this beam exhibited a maximum bending strength of 1540 lbs.  Testing different preform layups for strength qualities will be necessary to fully classify this composite and determine its viability for bridge construction.

Processing and Characterization of Nanotube-Reinforced Adhesives for in Situ Damage Sensing Applications

Zach R. Melrose, Limin Gao and Erik T. Thostenson
Department of Mechanical Engineering

It has been demonstrated that carbon nanotubes can be utilized as sensors for detecting the onset and accumulation of micro-scale damage in composite materials in situ. Future naval and aviation structures require joining of complex structural members. Research is aimed at extending the in situ sensing approach to hybrid composite joints to enable real-time damage sensing and health monitoring. This study focuses on processing and characterization of carbon nanotube (CNT) reinforced adhesives for applications in sensing of damage in adhesively-bonded joints. Experiments have been conducted on dispersing CNTs in a commercially-available high performance adhesive and determining the electrical properties and shear strength. The shear strength was determined by conducting a single lap shear test, where each specimen is composed of two composite substrates, and the electrical properties of the CNT are utilized to form a conductive network within the non-conductive adhesive. This network is destroyed during the test procedure and is witnessed in the increasing resistance of the joint which reflects the increasing damage. Each specimen was tested in tension until failure while simultaneous load and resistance data are collected for CNT/adhesive specimens containing 0.5 and 1.0 wt% carbon nanotubes. A new processing technique is currently being explored to decrease the resistivity of the CNT/adhesive joints for enhanced sensing response. 

Determining material properties of Homarus americanus exoskeleton and modeling of biomimetic laminate structures

Tom Mulrooney, Liang Cheng, Anette Karlsson, and James Glancey
Department of Mechanical Engineering

This project is an experiment with replicating the structure of arthropod exoskeletons, specifically that of the Homarus americanus (American lobster). On a microscopic level, it can be seen that the exoskeleton is composed of many layers of unidirectional, chitin fibers. The layers are stacked in a helix like pattern, where each layer is rotated a small degree from the previous. This pattern can be replicated using an S2-glass prepreg.  This current research focuses on two aspects of the project: determining the material properties of the lobster shell and modeling the composite using finite element analysis software. To find material properties, a three point bend test was chosen. To accommodate the small sample size a new test setup was designed and machined. The setup features an adjustable span length (10-40 mm at 10 mm increments), and rounded tips at the three contact points to prevent excessive indentation. A standard cast acrylic sample yielded accurate results for the elastic modulus test. To simulate the composite on the computer, ABAQUS FEA software was used. The tests were divided into long beam and short beam. Both were modeled with SC8R continuum shell elements. Each of these beams was dimensioned to match those that were being manufactured and tested over at CCM. When the data was compared, we found that the long beam simulations matched the experimental data very well. The short beam simulations were close, but some adjustments are being looked into to improve the accuracy.

Acoustic Characterization of Echogenic Liposomes

Daniel Russakow, Amit Katiyar, and Kausik Sarkar
Department of Mechanical Engineering

Liposomes are spherical volumes enclosed by a lipid bilayer, as are biological cells. They are likely to be more readily accepted by the body, and therefore a perfect candidate for biological applications. Echogenic liposomes are liposomes which also strongly reflect ultrasound, and so they can be used as contrast enhancing agents during diagnostic imaging. For successful clinical translation, they need to be accurately tested for their acoustic properties. This project focused on the design, development and analysis of an experimental setup that would allow such testing using milliliters of sample volume. After an acoustic setup was built, a well-researched contrast agent (Definity microbubble) was characterized using the new setup. The setup was concluded to be sufficient for liposome testing, which will commence shortly.

Destructive Test of a Steel Slab-on-Girder Bridge

Christopher A. Russo, Michael J. Chajes, Harry W. Shenton III, Jennifer McConnell, and Kervin Machaud
Department of Mechanical Engineering

This poster details the results of preliminary analysis, and the future planning, of a destructive bridge test scheduled to take place in the fall of 2009. Due to realignment and road improvements of roadways on the Delaware side of the Delaware Memorial Bridge, seven bridges have been scheduled for decommission by the Delaware River and Bay Authority. The University of Delaware has been given the unique opportunity to destructively test some of these bridges before decommissioning. The purpose of conducting the destructive bridge tests is to understand the behavior of bridges right up to their failure limit. One bridge has already been tested. The next bridge scheduled for testing and demolition is Bridge 7R. This bridge is located on the off ramp where Interstate 295 South merges into US13 South. As each subsequent bridge is tested, the test methods are refined and improved so that more precise data regarding bridge behavior can be collected. Due to the open window of time given from the test of the first bridge and Bridge 7R, which is scheduled for testing in the fall of 2009, the University of Delaware has been able to create finite element models to evaluate the expected failure modes, and to determine the locations for sensors to be placed on the bridge to yield the best data set for evaluating the bridge performance.  The models have also been used to evaluate the effect of skew on the ultimate capacity of bridges. The remainder of this poster will illustrate: (1) the preliminary models used and the results that they provided, (2) analysis the effects of bridge skew on ultimate capacity, and (3) a projected timeline for the test of Bridge 7R in the fall of 2009.

Ejector Design for Hydrogen Recirculation in the Fuel Cell Bus

Emmanuel Ungaro, Manish Bajpai, Suresh G. Advani, and Ajay K. Prasad
Center for Fuel Cell Research

In a Proton Exchange Membrane Fuel Cell (PEMFC), hydrogen and oxygen react electrochemically to generate power with water as the byproduct.  The University of Delaware’s fuel cell hybrid bus is equipped with a 20 kW PEM fuel cell stack to provide propulsion.  The fuel cell is supplied with hydrogen stored in roof-top tanks while the oxygen comes from the air.  The fuel cell is supplied with excess hydrogen in order to prevent fuel starvation at the electrode and to help with water management.  Hence, the exit gas stream from the anode of the fuel cell stack contains unused hydrogen which would be wasted if it were simply vented to the atmosphere.  In order to avoid fuel waste, methods have been developed to recirculate the hydrogen into the system.   The current method of circulating unused fuel back into the cell involves a pump that represents a substantial parasitic power loss (around 200 W).  This goal of this project is to design a fuel ejector that uses less power to recirculate the unused hydrogen back into the system.  The design process involves the use of a commercially available computational fluid dynamics package to ensure that the flow and pressure drop within the ejector are optimized for good performance.  Subsequently, the design will be fabricated and evaluated using a bench-top apparatus.  Once the design is proven to work reliably, it will be implemented in the bus.

Links: Summer 2009 Undergraduate Research Symposium, Symposium Abstracts from other Colleges and Departments,
2009 Undergraduate Research Summer Enrichment ProgramUnversity of Delaware Undergraduate Research Program, Howard Hughes Undergraduate Program.
Created  7 August 2009. Last up dated 18 August 2009 by Hal White
Copyright 2009, University of Delaware