I need a software capable of simulating the physics of a car going off a ramp. This is a stunt that I will be attempting and looks like the following:

A car, going off a curved ramp, doing a barrel roll, and landing on its wheels (right side up)

I have been provided with a model of the car that will be performing stunt, but I need to create the ramp. The following are the requirements needed for this stunt:

  1. The car must reach a peak altitude of 15 ft
  2. The car must do a single barrel roll while in the air
  3. The car must be going 150mph when it reaches the ramp

Before actually performing the stunt, i would like to simulate it to make sure my team has a high chance of not destroying the car.

I have tried to use Autodesk Maya, Blender, and Unity 3D to simulate this stunt and i have run into a problem: these programs use proprietary units of measurement (Speed is 27.02 instead of 150 mph) so i have no way of knowing what is the actual speed of the car (27.02 might convert to 137.8324 mph in reality).

My question is: Is there a program which can simulate 3D physics that have gravity, speed, and other vectors that could have inputs in imperial or metric formats?

  • $\begingroup$ Indeed, the programs you've tried are rendering programs, not simulators. (And game "physics" engines are in no way related to real physics.) Even if you were happy with their treatments of units, you should not trust their results to match the real world. $\endgroup$ – user10851 Aug 15 '15 at 0:59
  • $\begingroup$ Thank you for this information. Could you recommend a simulator program that might work for this project? $\endgroup$ – SuperMaku Aug 15 '15 at 1:01
  • $\begingroup$ Would BeamNG Drive work for this simulation? $\endgroup$ – SuperMaku Aug 15 '15 at 20:04
  • $\begingroup$ One of my co-workers did a simulation like that with simulationx: Flug ins Kirchendach The main consequence of the simulation is: Don't do it. The resulting trajectory is very parameter-sensitive. If you get the wrong wind you are killed. If your parameter measurements/estimations are wrong you get killed... $\endgroup$ – Tobias Dec 21 '16 at 15:15

Uhmm, yes, I do realize that this is probably too late, but my two cents are:


Simulating this problem for a car going 150mph with sufficient accuracy to ensure the driver doesn't get killed would be a major undertaking, which the original poster clearly would not be qualified for. I hope he is still alive and well...

But, just in case anyone sees this and gets inspired: You're putting your life on the line. At the very least, you would have to perform an experimental validation of your setup, perhaps using some sort of a dummy driver and using the exact same car.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.