The Auto Channel
The Largest Independent Automotive Research Resource
The Largest Independent Automotive Research Resource
Official Website of the New Car Buyer

PNGV Researchers to Deliver Toolkit for Designing Tomorrow's Vehicles

18 August 1998

PNGV Researchers to Deliver Toolkit for Designing Tomorrow's Vehicles PNGV researchers from USCAR, Southwest Research Institute (SwRI) of San Antonio, Texas, and TASC, Inc. of Madison Heights, Mich., are nearing delivery of their much-anticipated computer program that should speed the analysis and development of tomorrow's PNGV "supercar."  

The breadth of technologies being considered, along with the cost of research and development to build prototype cars of the future, requires that comprehensive analytical computer software be utilized to select and integrate the most appropriate technologies. Researchers hope to meet this challenge with the PNGV Systems Analysis Toolkit, a package of personal computer-based programs.  

While other simulation codes are available, PNGV researchers decided early on to start from scratch. Why? According to Dick Swiatek, a representative on the PNGV systems analysis technical team, "We started with a clean sheet of paper for several reasons. One is, because industry and government developed the code together, we could all agree on the technical assumptions in the code. This leads to a higher level of understanding for all of us who will be using the code. Our results will be more credible as well. Another reason is that other codes are 'backward-looking,' while our code is 'forward-looking.' Basically, this means that the PNGV Toolkit is a more realistic modeling approach that accurately represents the driver acceleration and braking inputs, much like driving an actual vehicle. Furthermore, we added a 'graphical user interface' that allows novice users to quickly learn how to use the code. This is important because people with a variety of technical backgrounds will be using the Toolkit." 

The Toolkit is used to evaluate powertrain configurations and components such as advanced heat engines and various energy storage devices. It provides an opportunity for component experts to see how their technologies perform in the whole vehicle system. 

"Engineers and researchers working on particular technological areas, such as materials, powertrains or electronics, need to be able to compare component options within their areas of expertise to the other components in the vehicles," said Mutasim Salman, chairman of the systems analysis technical team. "We designed the Toolkit to do this and more." 

Detailed performance predictions (miles per gallon, acceleration, emissions, etc.) for hybrid and conventional powertrains will be evaluated using the PNGV Toolkit. 

The Toolkit also allows researchers to conduct detailed studies of non-performance vehicle issues such as cost and reliability. The results of the studies include projection of component power and energy utilization, identification of areas of inefficiency, and evaluation of power and energy control strategies. These factors and others are considered in determining which components and systems would enable the midsize car of the future achieve up to 80 mpg at an affordable cost.  

The first version of the Toolkit was completed in December 1996. Researchers are continuing development of the Toolkit in a personal computer-based Windows environment.  

Three sections make up the Toolkit architecture. A "graphical user interface" facilitates vehicle and component definition, reports results, and analyzes cost and reliability. Vehicle models are created and evaluated in the MATLAB/SimulinkTM programming language through a program developed at SwRI. Finally, a database is provided for storage of component characteristics, vehicle configurations and simulation results.  

Vehicle analysis begins with the user selecting the vehicle configuration (e.g. conventional, or series or parallel hybrid. Hybrids have two power sources on board a vehicle - one that converts fuel into useable energy and another that lowers the demand placed on the first power source). Then, the powertrain components - including various engines, transmissions, motors, batteries, control strategies, etc. - are selected from the database library. A unique feature of the Toolkit is its capability to size the power requirements of the various components to provide the user with specified vehicle performance characteristics. The user then selects the simulation (vehicle maximum performance or the various emission and fuel economy drive cycles - city, highway, etc.). 

The actual simulation is performed using a "forward modeling approach." A driver model supplies a throttle or brake command to a power controller, with feedback to determine conformance to the desired vehicle speed and acceleration. Previous simulations have relied on a "backward modeling approach," where the desired speed and acceleration are used to compute the drivetrain energy losses and the required engine power. The forward approach allows more accurate representation of the system dynamics and control strategy, an important consideration in optimizing the complex hybrid vehicle power and energy storage system interactions. 

The vehicle performance, fuel economy and emission results are presented in either a summary spider chart format, or in detailed printouts. 

A simplified weight distribution effect on vehicle under/over steer characteristics is computed. It is based on the component weights required to achieve the desired vehicle performance and a pre-selected component placement scheme for front-wheel drive configurations. 

Vehicle and component reliability projections are based on component maturity and reliability growth model techniques utilized by the National Aeronautics and Space Administration (NASA). Reliability targets and the required test and development plans can be developed. 

The vehicle cost projections in this version of the software concentrate on the powerplant components. The component power levels, power density and cost-power relationships are used to compute the component cost. The cost of ownership projections include life cycle fuel cost as well as projections for component replacement, repair, insurance, residual value, etc. 

The software will provide the measurements necessary to establish future component development plans and targets required to achieve the vehicle program goals. 

This software is the result of a continuing team effort. The PNGV systems analysis team directs the development; SwRI developed the MATLAB-based performance and fuel economy module; and TASC, Inc. developed the graphical user interface and the handling, reliability and cost modules. In addition to Chrysler, Ford and GM team members, technical support is provided by Argonne National Laboratory, Idaho National Engineering Laboratory, Lawrence Livermore National Laboratory, National Renewable Energy Laboratory, Oak Ridge National Laboratory, Oakland University and the University of Michigan. Initial funding was provided by NASA and the U.S. Army National Automotive Center, with future funding support anticipated from the U.S. Department of Energy.