how accurate have you found it compared to being on the rollers ?
Depending on the algorithm and the accuracy of inputs, it can be very close (sometimes better, since it reflects REAL road whp)
First: Any roller is just trying to duplicate the real thing. So by definition, it is at best a close approximation.
Using time, speed and distance measurements will provide the best and most direct description of the vehicle dynamics, including acceleration (deceleration if you want), power and (wheel) torque.
Included are drivetrain losses (as with roller dynos), but also wind resistance, actual vehicle weight, etc.
The reasons this method is not the ONLY one being used by tuners, users, etc:
- vehicle is mobile (so all secondary test equipment has to be dragged along)
- test conditions are harder to control (climate, road surface)
- results are even less comparable from one vehicle to the next as with Dynos (which are all over the place as well).
