Physics demonstrations in BeamNG

This is a little experiment to show off the realism of BeamNG. I will calculate the mass of a car, and the static friction coeficient between the car's surface and the concrete ground, in two demonstrations.

The mass calculation is a bit redundant since the game already tells you the car's mass, but it's interesting to do anyway:

In this picture the car is perfectly steady, being held up with the node grabbing tool. The force required to keep it in the air with no acceleration is 10283 Newtons. So how do we turn this number into mass?
Well, if the car is stationary with no acceleration, that means the resulting force is null. Since the only forces acting on the car are its weight and the node grabber, we can conclude that the value of the car's weight (mass*gravitational acceleration) is equal to the force of gravity on the car. This means that the car's weight is also 10283 Newtons. Since weight is equal to mass*gravitational acceleration (m*g), g being roughly 9.8 m/s, we can conclude that the car's mass is 10283/9.8, which is equal to 1049kg.
If we check the in-game info on the car, we see the car's mass is 1050kg, which is very close to the calculated mass.


Now for the coeficient of static friction, I removed the front and rear suspension, as well as the exhaust, to make sure the only surface touching the ground is the car's body. We will also use the node grabber method for this experiment. But first, we need to lay down the proper calculations to make the whole process easier.

Force of static friction (Fs) is the force applied by the node grabber, and the value of Normal Force (N) is equal to the value of the car's Weight, given that on a vertical axis Weight and Normal Force are the only two forces being applied.
To obtain the value of Normal Force, we need to use the node grabber yet again:

The force required to keep the car stationary in the air is 6510 Newtons, which means the Weight and Normal Force are also 6510 Newtons.

The force required for the car to begin moving across the ground (Force of static friction) is 3291 Newtons. Using the calculation above, we get:
3291=Coeficient of static friction*6510
The resulting Coeficient of Static Friction is 0.56 aproximately.
In reality, the friction coefficient between steel and concrete is roughly 0.57 (Source: http://ascelibrary.org/doi/pdf/10.1061/(ASCE)0733-9445(1985)111:3(505)) , so the calculated result is once again very close to reality.
The minor differences between calculated results and real life values are purely a result of errors in my experimentation method (such as the car not being 100% steady or the force applied not being purely horizontal in the friction calculation)

In conclusion, the devs have made so much more than a "crash game". This is one of the best physics simulations out there, and the single best car simulator overall. The dev team's hard work has paid off amazingly, and will continue to for hopefully years to come.
Cheers

 

What other experiments/demonstrations can I do in BeamNG?

 

Aerodynamics/drag? Nice work.

 

Wish their air simulation was better.

 

i really wish the game would simulate lift better. at the moment it seems that stuff like this is not possible:



in my experience even when going 200mph over a ramp in a hillclimb (high downforce) car it will just fly straight.
if this was simulated in beamng the physics would be pretty much spot on.

 

i always knew beam was a lot more than a game with fancy crashes. every day it seems like i learn a new fact about this game that blows me away!

 

Hell, even slipstream isn't simulated in cars. Downforce and air drag is, just like engine power, made up by the devs.

 

Very well-calculated. I couldn't have shown it better myself. Excellent proof. I didn't even know you were capable of such show-and-tell, Slammington.

 

Try a tire physics demonstration, I've heard people from other simulators complain that the tire physics here are not good enough.

 

For anyone that wants more stuff like this can check out this thread:
http://www.beamng.com/threads/beamng-science.28426/
This thread by @Drivver generally talks about the science in beamng(as thread name suggests).
Specifically the thread talks about topics like the aerodynamics that beamng uses and centrifugal forces vs measured and actual speed and other stuff like that. It's an neat thread so check it out.
The thread also has some discussion about how beamng simulates the aerodynamics were even the staff puts their point about the aerodynamics.

 

This is the reason i love BeamNG.. damn. Good job all of you devs.

 

I am glad that you guys take such an obsessive approach to physics and vehicle systems. I remember back in the day that I got scolded by Lifter in the RoR development forums for even suggesting that instead of airspeed, the ABS system should be using wheel speed just like in real life. He just about ripped my head off for that one.

This is why I am beginning to like this game so much more (not that I didn't in the first place). Not that there is anything wrong with RoR, but it had/has a lot of little "cheats" built into the system to make it feel like it was doing things realistically. For example, speaking of aerodynamics, Fusedrag... it was like the complicated vehicle's aerodynamic loophole. Once Fusedrag was enabled in a vehicle, the sky was the limit in terms of speed.

I love BeamNG for the fact that when you come to a problem, rather than just implementing a cheat, you guys keep working until it is as physically accurate as the computer can handle. You don't just implement a silly work around, and if one is implemented, it is only there as a placeholder so that the game can function until a proper system is implemented, or is there as a simpler solution which is in place until enough optimizations are made to the physics engine to allow for the required computational power left over to run a more finely tuned and accurate system.

I applaud your approach to realism and honestly, being realistic as you guys are is the most difficult way of making a racing game. Like you said, the aim for the AI is to basically just have them controlling the vehicle just like a human and nothing more. No silly things to keep them on their line or to keep them from wrecking. It's all real. Honestly, I would put the word out to the IT departments of self driving car companies to develop their systems in your game. Nothing comes as close to a proper simulation as Beam does, and like you said, you are basically developing your own self driving car code in the game as we speak anyways!

 

AFAIK the game's approach to aerodynamics is not totally complete (as @estama said above) so it would not be possible to get the calculations to match up exactly with reality, correct me if I'm wrong.

Thanks, glad to hear it!

I could certainly try to verify a tire's coeficient of friction (of course there are many other aspects to tire simulation), for example. I believe the best tire/ground combo to do this with would be race slicks on dirt, as the game is a bit unreallistic in this aspect IMO (the slicks have amazing traction on dirt, in reality a car using slicks would lose control in a dirt road) once again please do correct me if I'm mistaken

Nothing short of absolutely amazing. I'm eager to see proper aerodynamics implemented (such as air turbulence simulation)

 

I dont need to hear anything else. Last i remember someone speaking about the subject was about how they weren't gonna try. I can be patient. I dont care how long it takes for multiple cars affecting each other's aerodynamics or ground effect to work as long as you actually do try to simulate it. You have all my support.

 

Im quite pleased with the drag shown here:

Original post can be found here:
http://www.beamng.com/threads/devel...r-contributing-post.1575/page-475#post-507276

 

That is downforce, not to be confused with drag ;p

 

Isn't it an mix of both that caused the issue shown?

 

As the person who basically implemented the Aero for that mod using LiftCoef instead of Dragcoef. I can tell you that the problem is simply due to the difficulty of reinforcing such a big piece of a jbeam that has to both break apart properly but also be able to give the vehicle the required downforce from the rear wing. If you go adding beams to make the wing not deform like the video above, it will not break properly, but without the beams it basically acts like a DRS system that is controlled by the speed.

 

Yeah it is, I was mistaken. Downforce is sorta a form of air drag so it's both.