Automotive Industry
By far, the main use of Mechanical Simulation products is in the automotive industry, mainly with CarSim. CarSim is a PC-based program for simulating the acceleration, braking, handling and ride of cars and car-trailer combinations over any road surface or terrain. A similar program — TruckSim — includes the dual tires and asymmetric steering systems needed to simulate medium or heavy trucks and the multiple vehicle configurations in common use. These products are the result of three decades of research on how to best simulate vehicle dynamics conducted at the University of Michigan Transportation Research Institute. Its validity has been well established by many automotive OEMs and Tier 1 suppliers where it is currently being used.
Product Design
Simulation has become an essential tool to engineers in the product development process. In the well-known “V-curve” representation of the process, CarSim is used as a predictive tool to refine system and component designs at the conceptual stage. Once hardware becomes available, CarSimRT (real-time) provides the environment for hardware-in-the-loop validation testing of components and systems.
Automotive manufacturers have recognized the value of math-based tools in improving the efficiency and timing of vehicle design and development. One manufacturer attributes savings of over $1 billion to adaptation of simulation tools.
Collaboration with Suppliers
Easy data exchange of CarSim information between engineers and managers, and between companies (OEM and suppliers) accelerates the communication process while minimizing errors. Vehicle parameters, advanced electronic chassis controls, driver maneuver descriptions, road information and the simulation results and animations can be shared among collaborators with a simple exchange of e-mail files. New control strategies for drive-by-wire systems, can even be downloaded using the web. With CarSim's versatile animator simulation results can be shared with people who do not have CarSim simulation software, but still need to advise or approve design and testing programs.
Development Testing
Test engineers can use vehicle simulation software to reduce development test time and cost. While some test procedures are relatively easy to perform, others are not. New procedures are constantly being developed to test electronically-controlled chassis systems. The driver inputs, sensor types, locations and data ranges, road surfaces, vehicle parameters and animated results can all be simulated in advance of testing. New chassis systems — including stability control, 4-wheel steering, and active suspension components — can be represented in CarSim or as Simulink models which are tightly integrated with CarSim.
The test engineer has the capability to perform planned tests in a simulated environment prior to more expensive experimentation on the proving ground. Since CarSim actually runs faster than real time, engineers can quickly preview the test procedure to establish sensitivity of vehicle and test site parameters to design a more efficient and robust test protocol. Using simulated test surfaces that can represent any road topology and friction characteristics allows the engineer to evaluate how unintended variables such as roughness, grades or friction variations will affect performance. Sensitivity to load variations in a given test can provide guidance for such small details as the need to control fuel tank fill level. Tire properties can be precisely controlled in the simulation world to identify the importance of tire wear as a test variable.
Road test results from actual experimental tests can be imported into CarSim allowing engineers to overlay simulated and actual vehicle results. This is an excellent tool for evaluating vehicle test data, sensor locations and measurement errors, or diagnosing problems.
Many test engineers have faced the problems of designing, installing and calibrating an instrumentation system without comprehensive knowledge of the nature and range of the vehicle responses to be anticipated. Out of range signals or low-level, noisy signal levels are factors that confound the measurement process and add unscheduled expense to testing programs. This is particularly true with new test procedures. A pre-test using simulation provides an effective picture of data ranges, how the data will be affected by test variables, and sensitivities to various vehicle parameters.
Test Services Providers
Why do contract test organizations need to add in-house vehicle dynamics simulation? Because it greatly improves the efficiency of their road testing capability. In a recent issue of Testing Technology International, Keith Read warned of the shakeout among testing organizations arising from cuts in OEM contracts. Test organizations are not unlike other automotive suppliers in that they need to assume a “lean and mean” posture to achieve success in the future. They must offer more full-service development and validation capabilities. Software programs such as CarSim provide test engineers with quick, reliable and powerful tools to evaluate new test methods, instrumentation, and vehicle chassis tuning sensitivities in lieu of days of test track experimentation. Dozens of simulated runs can be made in a few hours on a PC to help select the optimum variables for the final proving ground or road tests.
Testing organizations can significantly improve their value-added customer services using simulation results to communicate with product development and electronic control engineers at the start of new programs. Vehicle dynamics simulation adds to the insight and efficiency of vehicle road testing capabilities. The process of preparing and instrumenting a vehicle, selecting driver maneuvers, a suitable test road, and conducting the test is all replicated in CarSim in an easy to use and understandable manner. The graphical user interface of CarSim is recognized for allowing first-time users to run simulated tests after only an hour of self-guided experience. Over 100 example tests and nearly a dozen vehicle combinations allow engineers to “hit the ground running” in the simulation world.
Service organizations with an eye toward competitiveness — who want to offer more, faster, and cheaper — take advantages of the “virtual world” of vehicle simulation. The future belongs to these innovative companies.
An Example — NHTSA Fishhook Test
Simulation provides the ability to experiment with test procedures that are dangerous and difficult to control, or to compute performance variables that are hard (or impossible) to measure experimentally. This problem can be demonstrated, and is particularly relevant to the current focus on rollover resistance. Rollover behavior is highly nonlinear, dangerous, sensitive to transient dynamics, and subject to many variables in driver control. Variability in results from the multiple driver-controlled maneuvers that evaluate rollover resistance used around the world (Moose test, ISO lane change, Consumers Union lane change, etc.) argues the need for refinement in these test procedures. The National Highway Traffic Safety Administration in the U.S. has taken rollover testing to an even more difficult level with the “Fishhook” maneuver, which requires a robotic, programmable steering controller in place of the driver.
The “Roll Rate Feedback Fishhook” maneuver is quite complex. First the steering system is calibrated to find the steering wheel angle that produces 0.3 g of lateral acceleration. The steer angle for the fishhook maneuver is set at 6.5 times that angle, with a ramp speed of 720 deg/sec. At the desired test speed the vehicle is quickly ramped to this angle; then, when the roll rate drops to less than 1.5 degrees per second, it is suddenly ramped to the opposite steer limit and held.
The complexity of this fishhook maneuver makes it difficult to predict the vehicle behavior to be expected without the help of a vehicle simulation tool like CarSim. Furthermore, it is difficult to assess the significance of test variables (such as friction level of the skid pad or tire variables) on experimental results without the information available through simulation. Providing these extra capabilities, as a part of normal testing services, greatly adds to the value of a services organization.