Printable Version

Virtual-Vehicle Product Development

Auto teams can save valuable time using award-winning RAPTOR® software

By Scott McBroom

Scott McBroom is the manager of the Advanced Vehicle Technology Section, which produced the award winning RAPTOR program. RAPTOR was recognized as one of the top 100 inventions in 2004 by R&D Magazine. McBroom has led development efforts for simulation tools to improve the efficiency, safety and emissions levels for trucks as well as pursuing the evaluation, simulation, design and integration of electric, hybrid-electric and solar-powered vehicles.

The development of a new vehicle begins years before it hits the factory production lines. In a competitive market, new cars are released every year that are faster, more powerful and yet still fuel-efficient. For each vehicle platform, automotive engineering and design teams must work in parallel toward the same goal of making the vehicle concept into a reality. For major automotive manufacturers and suppliers, producing an all-new vehicle can be a tactical challenge that requires balancing technology, cost,logistics and timing.

Coordinating design ideas and new concepts into components such as the engine, transmission and body must be handled in an efficient, timely manner so that the final product can be tested and showcased within its original budget and schedule. 

A major problem facing original equipment manufacturers (OEMs) is coordinating the overall development process across a multitude of engineering teams, which are inevitably at different stages in the process. This requires the development, communication and verfication of design, test and manufacturing data, which ultimately describes the final product.

Southwest Research Institute (SwRI) has developed a commercial, off-the-shelf software program as an initial response to this need under the trademark of RAPTOR (Rapid Automotive Performance Simulator). RAPTOR is an application program written in Matlab®/Simulink® and allows powertrain components to be simulated in a virtual environment. Mathematical descriptions of vehicle powertrain components are assembled in software and simulated under various user-selectable driving schedules in a process of linking these powertrain components and sub-components from the engine to the wheels. RAPTOR allows automotive engineers then to analyze and optimize vehicle powertrain systems with regard to performance, efficiency and emissions production. This is a crucial step in powertrain product development. Co-developed with DaimlerChrysler in Auburn Hills, Mich., and recognized as one of the top 100 inventions in 2004 by R&D Magazine, the software was created for automotive and truck and bus manufacturers and their suppliers. 

RAPTOR makes the difference

As a software tool, RAPTOR facilitates the parallel computer-aided engineering (CAE) of vehicles. Parallel CAE consists of design, co-simulation, analysis and hardware-in-the-loop (HIL) development activities. Co-simulation allows engineers working in different software packages to use the same reference vehicle models in a common computing environment. For example, an engineer using a high-fidelity cooling system model, such as Flowmaster21 could evaluate the impact of a design change on vehicle performance in conjunction with an engine controls engineer who is using a high-fidelity engine model coded in Fortran or C++. In each case, both engineers could be using the common RAPTOR back-plane, once the interfaces are  developed. 

Another special feature RAPTOR offers is an integrated database that ensures configuration management of both models and data. Over the years, this has become a must for multi-user vehicle simulation activities. RAPTOR improves on competitive products primarily by incorporating a fully functional database for storing and synchronizing models, input data, simulation parameters and some key simulation results. Unlike competitive products, RAPTOR stores and manages both the data used for vehicle simulations and the Simulink models in the database. 

The current primary function of RAPTOR is to predict the performance of the vehicle and powertrain and to predict the loading that will result from operation over various drive cycles. This load information also can be used as input to the stress and fatigue analyses that are undertaken for the various powertrain components.

RAPTOR also can be used to develop control algorithms and calibration constants in a virtual environment long before complete hardware sets exist. This development involves exercising the modeled components and actuation hardware in the simulated environment and evaluating the response of the overall system. Calibration of engines and transmissions to the 65 percent level is the target for software-based control system development, and this likely will be obtained within the next three to five years. 

Modeling and simulation also can be used to improve the evaluation and development of powertrain components. This is termed HIL testing and can involve electronic controllers, engines, transmissions and other powertrain components. HIL testing of transmissions already has been demonstrated at SwRI. Engineers at the Institute are now enhancing RAPTOR to perform engine HIL simulation. Electronic controller HIL simulation is also an ongoing activity. SwRI has proven for its clients that HIL evaluation allows performance to be optimized prior to assembly in a vehicle, which measurably reduces the product development time and cost.

RAPTOR is being used to demonstrate engine HIL evaluation as a tool for engine development and electronic control unit calibration. This is a joint effort among SwRI, DaimlerChrysler-USA and A&D Corporation.

RAPTOR bridges the gap

SwRI's Advanced Vehicle Technology Section is undertaking a multi-year, focused effort called Virtual Vehicle Research and Development. RAPTOR is a large step in the concept that integrates every aspect of evaluation and development of vehicle design, analysis, evaluation and calibration. Ultimately, the goal behind the "model-to-vehicle" philosophy is to produce a vehicle that has been designed seamlessly in a virtual environment. This means that mixed-fidelity modeling is utilized in every facet of a vehicle system development cycle to predict and verify the operation of each subsystem before going into production. This would include both hardware functionality evaluation for power production, reliability and durability, as well as software performance for controllability, diagnostics and prognostics.

This approach incorporates a number of engineering disciplines, and it is absolutely indispensable to ensure that information for the design is consistent, accurate, understandable, useful and accessible to every member of the design team in a seamless manner. Not only must performance issues be targeted at the beginning but also reliability issues must be predicted at an early stage. Manufacturing issues must be considered during conceptual design and analytical simulations, and calibration processes should be integrated into the modeling and simulation environment. Finally, location no longer should pose a problem as design teams in different continents will be able to work in unison. 

Test data from actual vehicles can be combined with RAPTOR to automatically calibrate transmissions.

Conclusion

With RAPTOR, component models and control strategies may be created, interchanged and simulated to predict real-world results from a virtual environment. RAPTOR can be used to evaluate new technologies and perform trade-off studies comparing these emerging technologies to existing equipment. Individual component models developed using RAPTOR can be used in conjunction with hardware testing to simulate virtually any driving, ambient or operational condition. A major automotive OEM has validated and begun to integrate RAPTOR into its design process, which has already provided them reductions in costs and time. RAPTOR's potential continues to be explored beyond its initial applications, expanding its capabilities and applications to meet our clients' needs in different areas of vehicle design, development and calibration. v

Comments about this article? Contact Joe Redfield (210) 522-3729 or jredfield@swri.org.

Reference
[1] SAE Paper No. 2002-01-1208

Acknowledgments
The author gratefully acknowledges the contributions of the Vehicle System Research Director Gary Stecklein and the Advanced Vehicle Technology Section staff in the Engine, Emissions and Vehicle Research Division: Principal Engineer Joe Steiber, Program Manager John S. Bishop, Research Engineer Jack J. Harris, Research Engineer Gergory J. Ostrowski, Research Engineer Angela T. Trader, Group Leader Joe B. Redfield and Training, Simulation and Performance Improvement Division Analyst Theresa Huth.

RAPTOR real-world examples

  • RAPTOR currently is being used to develop engine HIL testing using a V6 gasoline engine and an AC dynamometer that replicates transmission transients and simulates vehicle loads on the engine. Hardware testing allows engineers to subject vehicle components to real-world vehicle loads without installing the component into a "mule" vehicle for testing, thereby reducing testing time and cost. 
  • The SwRI-managed Clean Diesel IV consortium, with members that include several manufacturers and suppliers of the diesel engine industry, used RAPTOR to support diesel engine design work. Using the wealth of data available from a major automotive manufacturer, RAPTOR provided performance comparisons and instantaneous power requirements for a particular class of vehicles - the extremely popular SUV.
  • Even with fuel-cell vehicle components not yet commercially available, a heavy-duty fuel cell-powered mining loader simulation was performed to determine fuel storage requirements and vehicle performance using RAPTOR and other SwRI-owned models. SwRI engineers navigated the highly configurable architecture of RAPTOR to design, develop and simulate the mining loader with minimal modifications to software. 
  • SwRI engineers recently used RAPTOR to quantify shift quality for an automated manual transmission in a small pick-up truck. The effects of shift "jerk" on the vehicle, and ultimately the driver, could be evaluated because of RAPTOR's high-fidelity transmission and shift scheduling models.
  • RAPTOR was used to baseline a class 8 tractor-trailer's performance. Baselining is often necessary to determine the efficiency improvements that can be obtained by replacing current technology components with advanced components or materials. In addition to obtaining baseline figures, engineers were able to predict torque requirements at the driving wheels to specify an equivalent electric motor drive system such that equivalent or better performance would be obtained from the vehicle in a hybrid-electric configuration.
  • An automotive manufacturer is using RAPTOR to support drivetrain engineering design decisions and to estimate corporate average fuel economy (CAFE) numbers for the Environmental Protection Agency (EPA). New vehicle models, engine and drivetrain options and component configurations are regularly evaluated during the overall platform development effort based on their impact to performance and fuel economy. Business decisions on vehicle model offerings are based on the simulation results produced by the automotive engineer and RAPTOR. During the first half of 2004, this automotive manufacturer used RAPTOR for more than 300 simulation projects and 19 major product reviews. Fourteen of the product reviews resulted in go or no-go decision points for the product.

Published in the Spring 2005 issue of Technology Today®, published by Southwest Research Institute. For more information, contact Joe Fohn.

Spring 2005 Technology Today
SwRI Publications SwRI Home