Penn Engineering
The School of Engineering and Applied Science, established in 1852, is composed of six academic departments and numerous interdisciplinary centers, institutes, and laboratories. At Penn Engineering, we are preparing the next generation of innovative engineers, entrepreneurs and leaders. Our unique culture of cooperation and teamwork, emphasis on research, and dedicated faculty advisors who teach as well as mentor, provide the ideal environment for the intellectual growth and development of well-rounded global citizens.
- Department of Bioengineering
- Department of Chemical & Biomolecular Engineering
- Department of Computer & Information Science
- Department of Electrical & Systems Engineering
- Department of Materials Science & Engineering
- Department of Mechanical Engineering & Applied Mechanics
- Energy Research Group
- Engineering Documents
- General Robotics, Automation, Sensing and Perception Laboratory
- Institute for Medicine and Engineering
- Kod*lab
- Laboratory for Research on the Structure of Matter
Search results
Dataset Disorder Enhances the Fracture Toughness of Two-Dimensional Mechanical Metamaterials(2025-01-20) Fulco, SageMechanical metamaterials with engineered failure properties typically rely on periodic unit cell geometries or bespoke microstructures to achieve their unique properties. We demonstrate that intelligent use of disorder in metamaterials leads to distributed damage during failure, resulting in enhanced fracture toughness with minimal losses of strength. Toughness depends on the level of disorder, not a specific geometry, and the confined lattices studied exhibit a maximum toughness enhancement at an optimal level of disorder. A mechanics model that relates disorder to toughness without knowledge of the crack path is presented. The model is verified through finite element simulations and experiments utilizing photoelasticity to visualize damage during failure. At the optimal level of disorder, the toughness is more than 2.6 times of an ordered lattice of equivalent density.Publication Supplementary Materials: Algorithmic Design of Kinematic Trees Based on CSC Dubins Planning for Link Shapes(2024-10-08) Feshbach, Daniel Adam; Chen, Wei-Hsi; Xu, Ling; Schaumburg, Emil; Huang, Isabella; Sung, CynthiaSupplementary materials for the paper "Algorithmic Design of Kinematic Trees Based on CSC Dubins Planning for Link Shapes", presented at The 16th International Workshop on the Algorithmic Foundations of Robotics (WAFR), 2024. Paper abstract: Computational tools for robot design require algorithms moving between several layers of abstraction including task, morphology, kinematics, mechanism shapes, and actuation. In this paper we give a linear-time algorithm mapping from kinematics to mechanism shape for tree-structured linkages. Specifically, we take as input a tree whose nodes are axes of motion (lines which joints rotate about or translate along) along with types and sizes for joints on these axes, and a radius r for a tubular bound on the link shapes. Our algorithm outputs the geometry for a kinematic tree instantiating these specifications such that the neutral configuration has no self-intersection. The algorithm approach is based on understanding the mechanism design problem as a planning problem for link shapes, and arranging the joints along their axes of motion to be appropriately spaced and oriented such that feasible, non-intersecting paths exist linking them. Since link bending is restricted by its tubular radius, this is a Dubins planning problem, and to prove the correctness of our algorithm we also prove a theorem about Dubins paths: if two point-direction pairs are separated by a plane at least 2r from each, and the directions each have non-negative dot product with the plane normal, then they are connected by a radius-r CSC Dubins path with turn angles <= pi. We implement our design algorithm in code and provide a 3D printed example of a tubular kinematic tree. The results provide an existence proof of tubular-shaped kinematic trees implementing given axes of motion, and could be used as a starting point for further optimization in an automated or algorithm-assisted robot design system.Publication Shape Recovery and Segmentation with Deformable Part Models(1987) Solina, FrancA method for recovery of compact volumetric models for shape representation and segmentation in computer vision is introduced. The models are superquadrics with parametric deformations (bending, tapering, and cavity deformation). The input for the model recovery is three-dimensional range points. We define an energy or cost function whose value depends on the distance of points from the model's surface and on the overall size of the model. Model recovery is formulated as a least-squares minimization of the cost function for all range points belonging to a single part. The initial estimate required for minimization is the rough position, orientation and size of the object. During the iterative gradient descent minimization process, all model parameters are adjusted simultaneously, recovering position, orientation, size and shape of the model, such that most of the given range points lie close to the model's surface. Because of the ambiguity of superquadric models, the same shape can be described with different sets of parameters. A specific solution among several acceptable solutions, which are all minima in the parameter space, can be reached by constraining the search to a part of the parameter space. The many shallow local minima in the parameter space are avoided as a solution by using a stochastic technique during minimization. Segmentation is defined as a description of objects or scenes in terms of the adopted shape vocabulary. Model recovery of an object consisting of several parts starts by computing the rough position, orientation and size of the whole object. By allowing a variable number of range points in a model, a model can actively search for a better fit (by compressing itself and expanding) resulting in a subdivision of the object into a model representing the largest part of the object and points belonging to the rest of the scene. Using the same method, the remaining points can be recursively subdivided into parts each represented with a single compact volumetric model. Results using real range data show that the recovered models are stable and that the recovery procedure is fast.Publication Kinegami: Open-source Software for Creating Kinematic Chains from Tubular Origami(Springer, 2024-07-16) Feshbach, Daniel AdamArms, legs, and fingers of animals and robots are all examples of “kinematic chains" - mechanisms with sequences of joints connected by effectively rigid links. Lightweight kinematic chains can be manufactured quickly and cheaply by folding tubes. In recent work [Chen et al. 2023], we demonstrated that origami patterns for kinematic chains with arbitrary numbers of degrees of freedom can be constructed algorithmically from a minimal kinematic specification (axes that joints rotate about or translate along). The work was founded on a catalog of tubular crease patterns for revolute joints (rotation about an axis), prismatic joints (translation along an axis), and links, which compose to form the specified design. With this paper, we release an open-source python implementation of these patterns and algorithms. Users can specify kinematic chains as a sequence of degrees of freedom or by specific joint locations and orientations. Our software uses this information to construct a single crease pattern for the corresponding chain. The software also includes functions to move or delete joints in an existing chain and regenerate the connecting links, and a visualization tool so users can check that the chain can achieve their desired configurations. This paper provides a detailed guide to the code and its usage, including an explanation of our proposed representation for tubular crease patterns. We include a number of examples to illustrate the software’s capabilities and its potential for robot and mechanism design.Publication Re-programmable Matter by Folding: Magnetically Controlled Origami that Self-Folds, Self-Unfolds, and Self-Reconfigures On-Demand(Springer, 2024-07-24) Unger, Gabriel; Sung, CynthiaWe present a reprogrammable matter system that changes shape in a controllable manner in real-time and on-demand. The system uses origami inspired fabrication for self-assembly and repeated self-reconfiguration. By writing a magnetic program onto a thin laminate and applying an external magnetic field, we control the sheet to self-fold. The magnetic program can be written at millimeter resolution over hundreds of programming cycles and folding steps. We demonstrate how the same sheet can fold and unfold into multiple shapes using a fully automated program-and-fold process. Finally, we demonstrate how electronic components can be incorporated to produce functional structures such as a foldable display. The system has advantages over existing programmable matter systems in its versatility and ability to support potentially any folding sequence.Publication Tritium Removal from CANDU Reactors(2024-06-18) Siefken, Ella; Lu, Jason; Ngo, PhuongWithin the next 50 years, the global demand for tritium will increase with the startup of fusion reactors, while the global supply (currently estimated at 25 kg) slowly dwindles due to its short half-life. Heavy water moderator from CANDU reactors is the only large scale source of tritium that can meet the rising demand. Only two tritium removal facilities are in operation worldwide, while the development of a third is uncertain. Currently, at the Bruce Nuclear Generation Station in Ontario, Canada, the tritiated heavy water is stored and transported to the Darlington Tritium Removal Facility, posing logistical challenges and risks of radioactive exposure during storage and transport. We propose the implementation of a tritium removal facility at the Bruce Nuclear Generation Station to continuously extract tritium from heavy water moderator from the four Bruce C units currently in the early stages of development. The heavy water would undergo electrolysis before being cooled to cryogenic temperatures of about 26–27 K and fed through a cryogenic distillation cascade that would produce gaseous tritium of 99.9% purity at a rate of 178 grams per year. Continuous tritium extraction would also maintain the radioactivity of the CANDU reactor moderator under 10 Ci/kg, an essential benchmark for reactor and environmental safety. Compared to existing tritium removal facilities, this process presents three novel advantages: 1) Direct electrolysis pre-treatment that foregoes the use of complex catalysts; 2) A thermally linked design that utilizes helium refrigerant to provide heating and cooling duty; 3) Optimized cryogenic distillation system that provides tritium product of higher purity that other similar processes. Assuming a tritium sale price of $30,000 per gram and a plant lifetime of 35 years, the tritium removal process presented is not profitable, with an ROI of -6.13% in the third production year and a negative IRR. However, this process design is highly valuable as tritium prices are expected to surge in the next decade as fusion plants reach technological readiness and require tritium to fuel fusion reactors.Publication Production of ATJ-SPK From Ethanol Feedstock(2024-06-18) Sheldon, Jacob; Stosich, Corbyn; Walker , MitchIn 2021, the U.S. government released the Sustainable Aviation Fuel Grand Challenge, which pledges a goal of supplying sustainable aviation fuel (SAF) to meet 100% of fuel demand by 2050. SAF currently makes up less than 0.1% of the total jet fuel industry and is nearly twice as expensive as jet fuel sold from a typical refinery (FAA, 2022). There are currently nine SAF production pathways that have been approved by ASTM, one of which is known as Alcohol-to-Jet-Synthetic Paraffinic Kerosene (ATJ-SPK). A 2019 white paper by Gevo outlines the principles of ATJ-SPK, where starchy alcohols are converted to isobutanol, which is then converted to paraffinic kerosene through well-established processes of dehydration, oligomerization, and hydrogenation (Gevo, 2019). An ATJ-SPK plant was designed with the intention of exploring the environmental and economic viability of a pure ethanol feed. To date, most developed ATJ-SPK plants have an isobutanol feed. The designed plant follows the three established steps: dehydration, oligomerization, and hydrogenation. For ethanol dehydration, Ni-HZSM-5 catalyst was used to convert ethanol to ethylene. The oligomerization and hydrogenation steps were both accomplished using two reactors in series with two unique catalysts. This design decision was made to target reactions in the C9-C16 range, ideal for kerosene jet fuel. For the first and second oligomerization reactors, Ni-H-β and Al2O3/SiO3 catalysts were used, respectively. Feed to the hydrogenation reactors consisted of mostly C8-C17 olefins, where a Ni-C catalyst was used in the first reactor, and a 0.3% Pt/Al2O3 was used in the second reactor. Paraffins were separated by size into SAF, diesel, and gasoline. Analysis revealed the economic viability of the designed ethanol feed ATJ-SPK process is highly dependent on the Sustainable Aviation Fuel Credit created by the Inflation Reduction Act. The credit gives $1.25 per gallon of SAF sold, given that the SAF has a 50% reduction in lifetime greenhouse gas emissions. The following process reduces GHG emissions by almost exactly 50%, putting the process at risk for not receiving the SAF tax credit if an unexpected source of emissions is discovered. Furthermore, the SAF tax credit expires in 2025, so there is little long-term economic viability of the designed process. With the discontinuation of government incentives, achieving the ambitious 2050 emissions goal for the aviation industry becomes even more challenging. If the United States is committed to these targets, additional tax incentives will likely be essential. The following design can be used as a baseline for future ethanol feed ATJ-SPK processes, and may prove viable if there are additional economic incentives established for SAFs as the 2030, 2040, and 2050 goals of the aviation industry to reduce emissions become a top environmental priority.Publication Hydrogen Production from Efficient Two-Step Water Splitting(2024-06-18) Bagchi, Rohan; Ghosh, Andreas; Murphy, KaraThe demand for renewable energy sources such as hydrogen is projected to increase in the next few decades as the world turns its sights towards reducing the effects of climate change. Hydrogen has recently been considered for use in the automobile industry as a power source for fuel cell vehicles because of its high energy density by mass. The greenest form of hydrogen production is through water electrolysis. Traditional water electrolysis, however, requires a membrane, which lowers efficiency and raises costs and safety risks. In this report, we design a process for two-step splitting of water by rotating cycles of electrochemical production of hydrogen and thermochemical production of oxygen without the use of a membrane. The process produces 28,000 U.S. tons of hydrogen per year with a co-product of 222,000 U.S. tons of oxygen per year. The electricity to power the electrolysis and other process units is sourced from solar energy. With a selling price of $1.02/lb of hydrogen - based on current prices for grey hydrogen, a selling price of $0.04/lb of oxygen, and a tax credit of $1.46/lb for the production of green hydrogen, the plant would achieve a return on investment of -1.87%, an internal rate of return of 34%, and a net present value of $152 million. In the best-case scenario where oxygen can be sold at a higher price of $0.30 for medical uses, the plant becomes much more profitable with an IRR of 67% and an ROI of 44%. The process’ voltage efficiency of 91.0% and HHV efficiency of 81.3% make it competitive with the best electrolysis technologies used in industry. Overall, the process provides one pathway towards a large-scale hydrogen economy.Publication Food Production Without Photosynthesis(2024-06-18) Dobkin, Gabrielle; Ellman, Sydney; Jacobs, CiaraWhile single-carbon molecules are not limited, the forms of edible carbon are. The rate at which photosynthesis converts carbon dioxide to biomass is limited by sunlight, climate, and water. With a global population of 8.1 billion and growing, it is imperative that food production remain secure despite changing and unpredictable conditions. Thus, single cell protein (SCP) production via industrial-scale fermentation proves to be an attractive avenue for producing food within a closed system using significantly less arable land and water. SCP is protein-rich biomass derived from unicellular microorganisms, such as yeast, bacteria, or algae. This report assesses the economic feasibility of producing human food grade SCP biomass using fermentation of the methanotrophic bacteria Methyloccocus Capsulatus with two different single-carbon sources as feed. This process was initially designed to produce 50,000 US tons of SCP biomass a year, it demonstrated capacity to produce upwards of 62,000 US tons/year. Therefore, this number was used in the final profitability analysis. The proposed location for the plant is Groves, Texas. The product was determined to have an ideal selling price of $7/kg. The methane-fed process was determined to be economically unfeasible after raw material costs and equipment costs were calculated. However, the methanol-fed process was determined to be viable after considering raw materials, equipment, utilities and other expenses. The methanol-fed process has an IRR of 29% and an ROI of 28%. With the fractional land requirements compared to traditional agriculture, the methanol-fed SCP production process proves to be worth pursuing based on environmental factors.Publication Post-Combustion CO2 Capture using Desublimation Technology(2024-06-18) Piotrzkowski, Kathleen; Izere, Dorine; Yang, BaileyCarbon dioxide levels in the atmosphere have risen dramatically over the past century, causing serious environmental concerns. The largest contributor to carbon emissions is the burning of fossil fuels for energy production. Much research is being conducted to develop new alternative fuels that do not release carbon dioxide, including hydrogen, solar, and geothermal. Despite these efforts, carbon-emitting fuel sources still supply about 80% of the world’s energy. To decrease carbon emissions in the current energy landscape, carbon capture is essential. Carbon capture selectively captures CO2 from the atmosphere through direct air capture (DAC) or point-source capture from flue gas streams. Existing carbon capture technologies only capture 0.4% of total emissions in the U.S., providing much demand and opportunity for the rapid development of new technologies. Carbon capture using desublimation selectively captures CO2 by decreasing the temperature inducing a phase change of CO2 from vapor to solid, producing pure carbon dioxide. This project proposes a process to capture 100,000 tons per year of carbon dioxide at 99% purity from a typical natural gas-fired power plant feed stream using desublimation technology. Our process offers a competitive design for carbon capture from diluted flue gas streams at $119 per ton of CO2.