ARC Collaborative Research Seminar Series
Winter 2018

ARC seminars are free and open to the general public. Center members can download the presentation files on our password-access online portal iARC. Non-ARC attendees please email arc-event-inquiries@umich.edu with your requests.

Parking & directions inquires: Contact arc-event-inquiries@umich.edu by 2:00 p.m. the day before the seminar.

Remote attendance via tele/video conference: Contact William Lim choonhun@umich.edu.

Refreshments will be served 9:15-9:30am. The talks will begin at 9:30 a.m. sharp.
Please note that event venue alternates between University of Michigan (Ann Arbor) and U.S. Army TARDEC (Warren)


February 9, Friday (9:30a.m. - 11a.m.)
University of Michigan, North Campus, GG Brown Building, room 2540 (Grand Conference Room)

1. Computational Discovery of Materials for Energy Storage: High Throughput Screening and Machine Learning
Dr. Donald Siegel, Assoc. Prof. Mechanical Engineering, University of Michigan

Abstract & Biography         Computational methods are now routinely used to characterize materials. A more impactful role for computation, however, lies in the accelerated discovery of new or overlooked materials with superior properties. This seminar will provide summarize our recent efforts targeting the discovery of materials for the storage of gaseous fuels such as hydrogen, and for the storage of thermal energy. High-throughput screening is used to assess tens of thousands of metal-organic frameworks (MOFs) for their ability to store H2 at high gravimetric and volumetric densities. These calculations identified several MOFs that have been experimentally-demonstrated to out-perform state-of-the-art adsorbents. Machine learning models are trained on the resulting MOF database, and will enable predictions on nearly 500,000 materials. In the case of thermal energy storage (TES), we describe an exhaustive first-principles computational search for promising TES materials based on hydrates and hydroxides mined from the Inorganic Crystal Structure Database. Several hundred hydration reactions were characterized with respect to their thermodynamics, energy densities, and operating temperatures. More than half of these reactions appear to be new to the TES literature. Promising reactions were identified for three temperature ranges: low (< 100°C), medium (100°C - 300°C), and high (> 300°C). Correlations linking TES performance with dozens of chemical features were quantified using a Pearson correlation matrix. These analyses reveal property-performance relationships, which have been used to formulate design rules for hydration-based TES systems.
 
        Don Siegel is an Associate Professor in the Departments of Mechanical Engineering, Materials Science & Engineering, and Applied Physics at the University of Michigan. His research targets the development of energy storage materials and lightweight alloys, primarily for applications in the transportation sector. Prior to joining UM, he was a Technical Expert at Ford Research and Advanced Engineering. Don has co-authored more than 70 publications, delivered approximately 80 invited lectures, and has been awarded several patents related to energy storage. He is a recipient of the NSF Career Award, the SAE Teetor Educational Award, and an NAE Gilbreth Lectureship. Prof. Siegel received a Ph.D. from the University of Illinois at Urbana-Champaign, with postdoctoral research performed at Sandia National Laboratories and at the U.S. Naval Research Lab. During the 2015-2016 academic year he was a VELUX Visiting Professor in the Department of Energy Conversion and Storage at the Technical University of Denmark.

2. Fatigue Resistance Optimization of Armored Vehicle Structures Using Weld Master S-N Curve
Dr. Nickolas Vlahopoulos and Dr. Pingsha Dong, Professors of Naval Architecture and Marine Engineering, University of Michigan

Abstract & Biography         As discussed in TARDEC’s 30-year strategy, TARDEC has been supporting Army readiness and has a continued focus on system sustainment and lightweighting of the vehicle structure. Producing vehicles that reliably meet structural design lives reduces the need for repairs and increases their availability for combat operations. In armored vehicles, structures fatigue life is dominated by the welded joints that typically have various complex geometric configurations with different material combinations due to ballistic performance and structural lightweighting requirements. To date, fatigue design and life evaluation methods for design evaluation of armored vehicles have been empirical at simple joint level under simple loading conditions, e.g., weld category approach by AWS and AASHTO and unable to take advantage of rapid advances in finite element structural modeling methods, resulting in significant uncertainties in structural lives for armored vehicles in operation. With recent rapid advances in mesh-insensitive structural stress method and master S-N curve approach adopted by ASME Codes and Standards, a quantitative evaluation of armored vehicle structural life becomes possible once fatigue behaviors in thick armor plate weldments are understood in the context of relevant material combinations and welding conditions.
        Research will be performed towards the following objectives:
• Developing a fundamental understanding of unique fatigue behaviors associated with thick plate joints used in armored vehicle structures through computational modeling and selected laboratory testing
• Establish a theoretical framework for data transferability of fatigue test data from different joint types, loading modes, thicknesses, and material combinations relevant to applications in armored vehicles, i.e., master S-N curve representation of joint fatigue resistance
• Develop computational algorithms for incorporating the master S-N curve developed for structural life evaluation and weight optimization for armored vehicle structures. The new algorithms will consider the candidate materials, the panel thicknesses, the layout of the welds, and the type of the welds as design variables. Improvement in performance metrics associated with lightweighting and reducing the manufacturing cost will be pursued, while meeting expectations for the fatigue life, and the survivability to loads from blast and ballistic impact.
        After an overview of the Army needs and of the technical foundation, our research approach to addressing both scientific understanding on dominant fatigue damage mechanisms in armor material weldments and robust computational modeling procedure suitable for various joint types will be discussed. Some past examples will be used to illustrate the validity of our proposed approach.
 
        Nick Vlahopoulos is a Professor in the Department of Naval Architecture and Marine Engineering (NAME). He has graduated 19 Ph.D. students, published 77 journal papers, presented over 100 conference papers, and authored chapters in two books. The areas of his research are: numerical methods in structural-acoustics, design of complex systems, and blast event simulations. Technology that he developed for mid-to-high frequency vibro-acoustic analysis is utilized by the US Naval shipbuilding companies for signature simulations. He also worked in formulating and implementing the multidisciplinary design optimization capability of the Rapid Ship Design Environment which is developed by the Naval Surface Warfare Center, Carderock under the CREATE Ships program. He has received funding from ONR, NSF, US Army, NASA, and Ford. He has served as the Undergraduate program Chair in NAME, and he has been the Graduate and Master’s program Chair in NAME for the past five years. He is an Associate Editor in the Journal of Acoustical Society of America. He has been collaborating with TARDEC since 1991.
 
        Dr. Pingsha Dong, Professor, Naval Architecture and Marine Engineering, and Mechanical Engineering (also as Director, Welded Structures Laboratory) at University of Michigan Ann Arbor, is the inventor of the mesh-insensitive structural stress method (also referred to as the Master S-N Curve Method) which has adopted by Ford Motor Company in FLOW since 2004. This method and others have been adopted by major national and international Codes and Standards, including the 2007 ASME Div 2 International Code, the Joint 2007 ASME FFS-1/API 579 RP-1 Fitness for Service Standards, mandated for design and analysis of pressure equipment and components by government agencies and regulators in over 50 countries worldwide. His research and teaching interests includes advanced fatigue and fracture assessment methodologies, computational methods for integrated manufacturing process simulation from manufacturability to structural performance evaluation. Prof. Dong has published more than 180 peer-reviewed papers in archive journals and major conference proceedings, including over 20 plenary/keynote lectures. He received a large number of national and international awards and recognitions, including: AWS 2015 Fellow Award, IIW (International Institute of Welding) 2014 Fellow Award, SNAME’s Elmer L. Hann Awards (both 2012 and 2007), IIW 2008 Evgenij Paton Prize, R&D Magazine’s 2006 R&D 100 Award for VerityTM development, TIME Magazine’s Math Innovator (2005), Aviation Week & Space Technology Magazine’s Aerospace Laurels 2004 Award, AWS 2004 R. D. Thomas Memorial Award (2004), SAE Henry Ford II Distinguished Award for Excellence in Automotive Engineering, 2003, ASME G.E.O. Widera Literature Award (2002), AWS 1998 Rene Wasserman Best Paper Award, 1998.

March 2, Friday (9:30a.m. - 11a.m.)
University of Michigan, North Campus, Duderstadt Center, room 1180

1. Remote Sensing Based Terrain Strength Characterization for the Next Generation NATO Reference Mobility Model Development
Dr. Thomas Oommen, Assoc. Prof. Geological Engineering, Michigan Technological University

Abstract & Biography         Determining the terrain strength properties is critical for achieving accurate mobility performance prediction as well as reliable operational planning using the NATO Reference Mobility Model (NRMM). The NRMM is a computer-based simulation tool that NATO forces utilize to predict the capability of a vehicle to move over specified terrain conditions. Even though NRMM is useful for the NATO forces, it has several significant limitations. One of the major weaknesses is that it relies heavily on in-situ soil measurements, which is difficult to obtain from unknown territories and combat zones. Therefore, having the ability to remotely sense the terrain strength properties of the areas that lack in-situ soil measurement and to utilize this remotely sensed data for NRMM model is critical for the mobility of forces in unknown territory. In this study, the relationship between soil strength and hyperspectral and thermal remote sensing are explored. Previous studies have shown that individually these sensing techniques are promising for strength characterization but conclusive validation of its applicability in different terrain conditions have not been verified. The objective is to combine the output from both the sensors to derive a robust estimate of terrain strength characteristics. The relationship between output obtained from the sensors will be analyzed using machine learning algorithm that has shown promise to derive complex nonlinear relationship.
 
        Thomas Oommen is an Associate Professor in the Department of Geological and Mining Engineering and Sciences at Michigan Technological University. His research interest is at the interface of geotechnical engineering, remote sensing, and machine learning. He has worked to resolve different geotechnical engineering problems that could benefit from the remote sensing dataset and the computing capabilities of machine learning and artificial intelligence. In the last 7 years, he has received over 6 million USD in funding to perform research for agencies such as World Bank, North Atlantic Treaty Organization (NATO), National Science Foundation (NSF), National Aeronautics and Space Administration (NASA), United States Department of Transportation (USDOT), United States Geological Survey (USGS) etc. He is the secretary of the engineering geology and site characterization committee of the American Society of Civil Engineers (ASCE) and the editorial board member of several journals. He has been invited to deliver keynote at several conferences for his work on applying remote sensing to solve geotechnical engineering problems.

2. A Decision-Based Mobility Model for Semi and Fully Autonomous Vehicles
Dr. Vijitashwa Pandey, Asst. Prof. Industrial & Systems Engineering, Oakland University

Abstract & Biography         This research is aimed at developing a formal methodology for evaluating ground vehicle systems and facilitating acquisition decisions pertaining to them. This seminar will outline our efforts in this area. A decision-theoretic approach will be presented which will begin with the development of a set of desiderata. The desiderata provide an axiomatic basis for the research work. For system evaluation and acquisition decision-making, a single unambiguous metric for mobility is generally essential. Fortunately, formal decision analysis ensures that such a metric, in the form of a utility function, exists. Careful attention will be given to the mobility metric of choice, attributes that affect it (including cost), decision variables and sources of uncertainty. To identify the attributes that affect mobility, one needs to outline what attributes the decision maker(s) has active preferences over. First, an exhaustive list is created which is then reduced to a reasonable number of attributes and then preference orders and ranges of negotiability are determined. Some examples in the context of mobility are speed made good (SMG), trafficable percent area (TPA) and cost of mobility among others. Existence of preferential and utility independence conditions also allows us to use the famed multilinear functional form, and will be discussed. Value of information (VOI) studies can be performed and results from our recent investigations in this area will be presented. VOI studies can answer questions such as - How much money should be expended in physical tests and simulations, when comparing two vehicles with uncertain mobility performance and/or under uncertain operating conditions. The seminar will also outline currently underway and planned simultaneous studies which will be performed using simulation environments such as ANVEL to help pose preference assessment questions and validate the recommendations.
 
        Vijitashwa Pandey is an assistant professor in the Industrial and Systems Engineering department at Oakland University in Michigan. He received his PhD from the University of Illinois at Urbana-Champaign in Systems Engineering. His research interests are in the areas of Engineering Design-Decision Making, Optimization, Reliability Engineering and Sustainability. He has authored two textbooks focused on engineering design-decision making as well as various peer-reviewed research articles in journals and conference proceedings. His work has received recognition in the form of best paper awards at ASME IDETC conferences, as well as the Arch T. Colwell Merit award at SAE World Congress.

April 6, Friday (9:30a.m. - 11a.m.)
U.S. Army TARDEC, 6501 E. 11 Mile Road, Warren, MI 48397-5000
Building 200B TARDEC University Class Rooms A&B

1. Robust Terrain Identification and Path Planning for Autonomous Ground Vehicles in Unstructured Environments
Dr. Jeremy Bos, Asst. Prof. Electrical Engineering, Michigan Technological University

Abstract & Biography         Autonomous Ground Vehicles (AGVs) operating in challenging off-road or unstructured environments require, at a minimum, two pieces of information. Specifically, the AGV must have or use knowledge of its own dynamics. It must also estimate the traversability of the local terrain. Using both pieces of information an AGV can, via any number of methods, plan trajectories by connecting traversable terrain segments. A goal of this work is the development of a robust three dimensional path planner for small AGVs that incorporates both vehicle dynamics and rich terrain information. We anticipate this local path planner will incorporate feedback from a terrain identification system developed by our commercial partner. This system is capable of estimating tire-terrain interaction based on wheel displacement and acoustic measurements. We also aim to extend the capability of this system to predict, rather than simply measure, tire-terrain interaction by incorporating information provided by onboard camera and LIDAR systems. Proposed path planning approaches will be tuned and evaluated in via Monte Carlo simulation and verified by extensive testing on well-characterized test courses. Year one of this effort will focus on a simple skid-steer wheeled AGV platform. This talk will cover the motivating works for this project and describe our approach, including candidate path planners, in detail. We will also discuss how this work may be expanded to larger, more complex vehicle architectures.
 
        Jeremy Bos is an Assistant Professor of Electrical and Computer Engineering at Michigan Technological University. His relevant research interests are is in what he calls “Autonomy at the End of the Earth” focusing on the operation of unmanned and autonomous ground vehicles off-road, in unstructured environments, and in inclement weather. Before starting his position at Michigan Tech Bos was a postdoctoral research fellow with the Air Force Research Laboratory under the National Research Councils Research Associateship program. At Michigan Tech Bos teaches graduate courses on Robotics and co-advises MTU’s entry into the General Motors/SAE AutoDrive Challenge.

2. Physics Augmented Artificial Intelligence for Automated Vehicles
Dr. Huei Peng, Roger L. McCarthy Prof. Mechanical Engineering, University of Michigan

Abstract & Biography         Perception, path planning and control are three important tasks in developing automated vehicles. For automated vehicle functions such as super-cruise (GM’s trademark) or Autopilot (Tesla’s trademark), the fundamental functions include lane/curvature detection, obstacle detection, path planning, and steering/speed control. Existing methods in the literature can be largely divided into two categories: end-to-end, and step-by-step. End-to-end approaches compute control actions directly, while in the step-by-step approaches, the tasks are separately conceived, designed, and validated. The step-by-step approach dominates the automotive field for many decades, until the new kid on the block, the artificial intelligence (AI), emerges, promising to totally revolutionize all aspects of the modern society, including design autonomous vehicles. AI concepts were promoted by “tech” companies such as Nvidia, Comma.ai, AutoX, and Argo.ai. Many of the end-to-end promoters believe it is not necessary to use vehicle dynamics/physics based knowledge and models—the final control decision can be figured out by the “intelligence” itself. A fundamental question this new ARC project aims to answer: is it true that dynamics/physics are yesterday’s knowledge and is becoming obsolete by the AI approach? Or, they still have some value in the design of autonomous vehicle control systems? If so, how do we use these two approaches in a complementary way? Is there a systematic way to identify tasks more suitable for AI vs. those more suitable to physics/dynamics?
 
        Huei Peng received his Ph.D. in Mechanical Engineering from the University of California, Berkeley in 1992. He is now the Roger L. McCarthy Professor of Mechanical Engineering at the University of Michigan. His research interests include adaptive control and optimal control, with emphasis on their applications to vehicular and transportation systems. His current research focuses include design and control of electrified vehicles, and connected/automated vehicles. In the last 10 years, he was involved in the design of several military and civilian concept vehicles, including FTTS, FMTV, Eaton/Fedex, and Super-HUMMWV—for both electric and hydraulic hybrid concepts. He served as the US Director of the DOE sponsored Clean Energy Research Center—Clean Vehicle Consortium, which supports more than 30 research projects related to the development of clean vehicles in US and China. He currently serves as the Director of Mcity, which studies connected and autonomous vehicle technologies and promotes their deployment. He has more than 250 technical publications, including 110 in referred journals and transactions and four books. The total number of citations to his work is more than 17,000. He believes in setting high expectation and helping students to exceed it by selecting forward-looking and high-impact research topics. One of his proudest achievements is that more than half of his Ph.D. students have each published at least one paper cited more than 100 times. Huei Peng has been an active member of the Society of Automotive Engineers (SAE) and the American Society of Mechanical Engineers (ASME). He is both an SAE fellow and an ASME Fellow. He is a ChangJiang Scholar at the Tsinghua University of China.

ARC members can download the presentation files on our password-access online portal iARC.
Non-ARC members please email arcweb-info@umich.edu with your requests.