Where can I find a dynamical model for a car with an automatic transmission?

Question

I am interested in simulating a dynamical model of a car to study aspects of "hypermiling" (minimum-fuel optimal control). In order to do this, I would like to use the longitudinal equations of motion for a conventional fuel-injected car with an automatic transmission. If possible, the model should take accelerator deflection as a control input, and the forward vehicle speed as the output, while also modeling the fuel consumption. Hopefully, it would be based on some combination of empirical data and mathematical principles. Are there any conference/journal papers, automotive engineering books, or other sources that would have such a model (and is available to the public)?

Thanks for reading!

Answer 1

Interesting question.

Such a simulation is certainly possible, but it would require knowing the complete control algorithm for a modern automatic transmission, and the fuel injection map for the engine control module.  I think most car companies would be rather unwilling to give you such proprietary information (you might take it to a competitor).

You could write your own algorithm, of course.  You'd need to research internal combustion to find out how much fuel a car needs as a function of airflow, engine load, and engine RPM.  Then you'd need to create a function that gave engine torque as a function of fuel input, throttle position, and engine RPM.  Then you'd need to model all the loads on the car, including driving torque, rolling resistance, and aerodynamic loads.

You'd also need logic to decide what gear to be in, as a function of throttle position and vehicle speed.

It would be quite a project, and you'd be guessing for some of the equations, so it wouldn't be exactly realistic, but with some research, I think you could get close.

Answer 2

I've seen fuel-spec fuel consumption- torque graphs in performance car magazines in their reviews. Blow it up on a photocopier and digitize. You might also be able to get more accurate data from the manufacturer.

Transmissions are a little more difficult. There are 3rd party performance trannies that publish better data. Your looking a few percent loss through the gear trains. Some report net power at wheels that can give you that. Toughest component to get will be the torque converter. Again, check 3rd party guys.

Another approach is to dyno a stock car. Many performance places have dynos. Pick a 2 wheel drive vehicle, 'cuz more dynos will handle them. This doesn't break it down between engine and rest of power train though.

Every car will be different; a decent model should be sufficient.

By the way the DOE (deparment of energy) also did lots of dyno studies in the 70's, probably still do some. You might be able to google some.

Answer 3

Automotive manufacturers might be a source. The government would most likely have something. Try NASA for such information since they do more than just model airplanes and space shuttles. There are other organizations that might have similar information. You'll have to include slip loss for the transmission, up to the stall speed of the torque converter. Using an automatic transmission is a bit of a lossy proposition because of the way those transmissions do operate. A manual transmission would producer better mpg. Vehicle design for aerodynamics needs to be considered. For what you want to do, would it not be more advantageous to look for better fuels, or alternative fuels that are easily obtained as opposed to just using current sources? There is just so much that can be done for the common internal combustion engine. There are specific laws that need to be considered for just how much energy is available for any given type of fuel.