ARC Collaborative Research Seminar Series
Winter 2013

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

If you wish to attend the seminar remotely, please contact William Lim (williamlim@umich.edu) for teleconference details.


February 22nd, Friday (9:15 - 11:00am)
University of Michigan, UM North Campus, Duderstadt 1180

On Ground Vehicle Battery Cooling

Bilevel Multiobjective Optimization for the Battery Thermal Packaging Design
Brian Dandurand, Paolo Guarneri, Georges Fadel and Margaret M. Wiecek, Clemson University

        The battery packaging design requires the optimization of the battery position in the vehicle subject to geometric constraints, and the simultaneous optimization of cell layout inside the battery pack while considering thermal aspects. Since the vehicle and battery design problems are typically addressed by separate teams and each design is driven by multiple performance criteria, the problem requires an approach to generating the Pareto set of the bilevel problem by coordinating computations based on the level-specific information. A block-coordinate-descent-type algorithm is proposed for the coordination of the two subproblems. When applied to the battery packaging problem, it converges to (local) Pareto designs of the bilevel problem. Tradeoff exploration on each level and negotiation between the levels drive the battery shape.

Electro-Thermal Cylindrical Battery Model for Robotic Packs
Youngki Kim, Shankar Mohan, Jason Siegel, Anna Stefanopoulou (University of Michigan); Yi Ding (GVPM, TARDEC) and Matt Castanier (Analytics, TARDEC); Dyche Anderson and Yonghua Li (Ford)

        Many of the military Packbot and Talon unmanned ground vehicles (UGVs) are powered by packs of cylindrical battery cells. These cells have lithium cobalt oxide electrodes that are sensitive to temperature, yet the packs are completely sealed and hence do not have any active cooling. With the goal of improving the performance and reliability of UGV batteries in the field, we have established an equivalent circuit electro-thermal model. The model accounts for the nonlinear dependency of the heat generation on temperature in healthy cells, which can be critical for avoiding damaging operating conditions and the onset of thermal runaway.


March 22nd, Friday (9:15 - 11:00am)
University of Michigan, UM North Campus, Duderstadt 1180

On Vehicle Electrical Energy Storage

Electro-Thermal Planar Dynamics and Control of Prismatic Li-ion Cells
Sun Ung Kim, Jason Siegel (co-PI), Anna Stefanopoulou (co-PI), Charles Monroe (PI), U. of Michigan

        Li-ion cells are becoming ubiquitous in civilian vehicle applications because of their large gravimetric and volumetric energy densities. A large-format prismatic cell configuration has advantages for high-power applications, but exhibits significant in-plane two-dimensional temperature variation. Since electrochemical properties within the cell depend on temperature, local temperature variations affect the local utilization of the electrodes during charge and discharge – an effect which is poorly understood. We will discuss how different control modes affect the temperature and reaction distribution within large-format prismatic cells operating at high rate. Our approach is both experimental and analytical, with an aim to develop models that efficiently calculate the most significant features of the temperature evolution within operating batteries. This seminar will focus on our development of streamlined analytical models to rationalize the observations of cell-level thermal response of large-format prismatic cells, as well as the implications of electro-thermal coupling for battery operation during cold-start or high-temperature operation.

High Energy Density Asymmetrical Capacitors
Priyanka Pande, Abdoulaye Djire, Prof. Paul Rasmussen, Prof. Levi Thompson (PI), U. of Michigan; Dr. Yi Ding, U.S. Army TARDEC; Dr. Saemin Choi, Inmatech, Inc.

        Batteries are the principal devices used for military and commercial energy storage applications. These devices can have energy densities exceeding 100 Wh/kg, but this energy is difficult to fully access in pulsed and high power applications. Supercapacitors offer much higher power densities and could complement batteries in pulsed power applications, however, their low energy densities are only sufficient for relatively short pulses (a few seconds). Our research is exploring the feasibility the using asymmetric cell designs and new, high capacity materials to produce asymmetric supercapacitors with energy densities that out-perform currently available devices and enable applications with longer pulses. This talk will describe our progress including a summary of performance characteristics for the materials and prototype cells.


April 12th, Friday (9:15 - 11:00am)
University of Michigan, UM North Campus, Duderstadt 1180

On Vehicle Dynamics and Control

Vehicle-Dynamics-Conscious Real-Time Hazard Avoidance in Autonomous Ground Vehicles
Jiechao Liu, Tulga Ersal, Jeffrey L. Stein (PI), University of Michigan; Paramsothy Jayakumar, James Overholt, TARDEC; Steve Rohde, Mitchell Rohde, Quantum Signal

        Unmanned ground vehicles (UGVs) are gaining importance and finding increased utility in both military and commercial applications. Even though the term UGV typically referred to small ground robots in the past, current efforts are also targeting larger size platforms. In addition, research does not focus on teleoperation only, but also on autonomy. However, due to size, speed and operating conditions, larger size platforms have significantly different dynamics than small robots, and therefore the existing hazard avoidance algorithms developed for small robots may not deliver the desired performance in larger platforms. In this talk we will present our efforts on developing a model predictive control (MPC) based hazard avoidance algorithm that is aware of the dynamic limitations of the vehicle and can thus push the vehicle to its limits to maximize its performance. To achieve this, higher fidelity models are needed to accurately predict the dynamic limitations of the vehicle. We will focus on the impact of model fidelity in MPC and present a comparative study with a lower and higher fidelity vehicle model employed in MPC.

Off-Road Soft Soil Tire Model Development, Validation, and Interface to Commercial Multibody Dynamics Software
Shahyar Taheri, Saied Taheri (co-PI), Corina Sandu (PI), Virginia Tech

        The dynamics of tire-terrain interaction plays a significant role in studying tire and vehicle performance. Tire dynamics is influenced by tire structure and tire-terrain interaction. The proposed model consists of three layers, representing different tire sections; each layer contains multiple lumped masses connected to each other with springs and dampers in various combinations. The tire interaction with the terrain is obtained using an innovative ground model, which adapts its boundary conditions based on the dynamics of the tire elements.
        In addition to model development, experimental testing has been done to supply data for model validation. Experimental work was performed using the indoor terramechanics test rig at AVDL that can control slip and normal load applied to an instrumented tire driving over a deformable soil bed. The tests provided substantial data for sinkage and all forces and moments caused by the tire-soil interaction under various conditions. The drawbar pull ratio versus slip ratio is one of the relations of particular interest.


April 19th, Friday (9:15 - 11:00am)
University of Michigan, UM North Campus, Duderstadt 1180

On Vehicle Design & Optimization

Preamble
Dr. Richard Gerth, U.S. Army National Automotive Center (NAC) Tank Automotive Research Development and Engineering Center (TARDEC)

Optimal Crowdsourcing Framework for Engineering Design
Alex Burnap, Max Yi Ren, Rich Gonzalez, Panos Papalambros (PI), U. of Michigan

        Crowdsourced evaluation is a promising method for evaluating attributes of design concepts that require human input. One obstacle to obtaining both accurate and comprehensive design evaluations is the signal to noise ratio of high ability to low ability participants within the crowd. In this paper we introduce a Bayesian network capable of finding participants with high design evaluation ability, so that their evaluations may be weighted more than those of the rest of the crowd. The Bayesian network also estimates a score of how well each design concept performs on each required attribute. Monte Carlo simulation studies were conducted to test the quality of the Bayesian network on a variety of crowds consisting of participants with different evaluation ability. The results suggest that the Bayesian network estimates design attribute performance scores much closer to their value than simply weighting the evaluations from all participants in the crowd equally. This finding holds true even when the subgroup of high ability participants is a small percentage of the entire crowd.


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.