what is this drag coefficient measured in? what is a normal value? what effect does a value of "35" have on something so small? if you take into account the rolling resistance issues affecting top speed that have been mentioned before, is that 10-20% figure still accurate?
The only thing I can add to this thread right now is "You shall see", when the Capri is finally finished, you'll all see how it is impossible to drive with no downforce, yet how easy it is with the wings on. Even the diffuser is functional, or should be, due to its coltri construction
Estama can answer the question about the drag coefficient, as to what unit it uses and how it works. All I can tell you is that it recreates real world behavior pretty well - the top speed reduction from aero drag is accurate, and lift/downforce are simulated realistically. The hundreds of hours I spent tuning the Bolide to make it more stable and better handling had absolutely nothing to do with aerodynamics and made a colossal difference in grip and stability, including grip while getting on the throttle mid-corner. It was simply suspension and tire tuning. It's extremely stable at the moment - so much so that I didn't even bother adding a wing, because it doesn't need it. But it's been proven that downforce does work if you build a spoiler properly.
None of the updated cars will work without the updated binaries (.exe and .dll files) which are currently missing a lot of features and have a lot of nasty bugs. So we really can't release anything right now.
IIRC, dragCoef 100 should be equivalent to a flat plate perpendicular to flow, which Wikipedia says has a drag coefficient of 1.28. The triangles don't have an effect on each other, there's no flow around the car or anything. Since a car has lower drag than a collection of flat plates, dragCoef of 10 to 50 for different parts of the car would be a good starting point. An airfoil would be more efficient than a flat plate, so I'd use at least 100 dragCoef for downforce or lift wings. My planes need a few hundred dragCoef just to stay in the air at a reasonable speed. I'm pretty sure there is still too much rolling resistance in every car. 0 aero drag should have a bigger effect on top speed. Once I did a test with Grand Marshal, put frictionless feet and a thruster on it, then calculated drag coefficient from the speed it reached. I think it was under 0.20? Similar tests could be done for other cars too, or simpler shapes.
In physics, the drag coef doesn't have a unit, it isn't a dimensionless quantity. Concerning BeamNG's drag coef, i didn't want for it to "look" like a real drag coef. This is why i scaled it by 100, so it'll look more like a percentage rather than like a drag coef that you would find in real life. The reason for that is that drag in our triangles is double sided. Think of a car's cabin (ascii art ahead): _,-----. <- o-----o The "." at the rear of the car won't be "exposed" to the main air flow in real life, whereas in BeamNG it'll be exposed to it [1]. So in BeamNG you'll need to calibrate the drag coef of the "." triangle, taking this "exposure" under consideration (you'll have to lower the drag coef of it). So, to summarize, BeamNG's drag coef has some peculiarities, that differentiate it from the real life drag. Due to that, it is by design made to look like a percentage (it is roughly equal to real drag coef * 100), to make it obvious that it is not similar to the real life drag coef. [1] If we had proper air flow simulation, the air would be simulated going around the cabin, so the "." would not be directly exposed to the simulated air. We are still thinking of ways to do the full air flow simulation in real time. BTW, deject3d, i liked your investigation very much. Thank you for sharing it .
thanks for the response. i can't really say i understand the effects of a spoiler at arbitrary drag coefficients still, but thanks nonetheless. i think it would be valuable in the future if there were slightly more telemetry options for measuring some of the finer data points that a car is subjected to. a wind-tunnel type test (would be awesome to show downforce measured in newtons spread over the car), more readable wheel friction data, current suspension travel + limits, the weight distribution of a car, etc. just keep it in mind
Concerning a spoiler, it is always exposed to the air flow, so you could use in BeamNG, a real spoiler's drag coef * 100. As for the telemetry options, we've added a per wheel weight distribution telemetry option. It doesn't have a GUI yet, but the per wheel sensor readings are collected.
I know you have a ton or work to do but you should really post here more, your posts are always so informative there always worth reading.
lol he is not exaggerating we are thhe players we know , your a develpoer so fix it cause its 100% true this game just slides everywhere so FIX IT
Wait I gotta go to the store first. Gotta floor the throttle, take turns with my steering maxed, and wonder why I had a crash. BeamNG is pretty good in its tire model if you treat it right. And aren't using crap mod cars.
Yes, because his entire fucking existence revolves around you. Look at the date of the post above yours, yes, you bumped a 6 month old thread demanding that your needs be catered to. Fuck off.
When ever I use the supercar or any car that goes above about 40 mph on road surface, it feels like I am driving on ice. The car bounces horribly, when I use the brakes the car comes even more unstable, it drifts out of control. The cars have no traction control! It feels unrealistic, mostly because I have experience from driving cars like that in real life. The cars in the game respond slowly, even if I'm going like 10 mph around a curve the car spins out of control. Very unrealistic. I was hoping for a game that would be easy to control. Instead, I got a game where it is a struggle to keep the car balanced and stable.
The bolide is a mid engined rear wheel drive supercar from the 80s, traction control wasnt a thing back then, ABS wasnt standard even on supercars back then (bolide also lacks this). Feather the brakes, feather the throttle, it becomes much more controllable. Real cars you don't turn lock to lock, however BeamNG doesnt limit the steering input with speeds (yet) so full steering input will put the car to full lock too. These points all combine to leave the bolide as a very temperamental car, however with some finesse on a gamepad or better still an actual steering wheel, it can be tamed. It will still want to kill you, but you can make it your bitch, it has dropped its soap in the prison shower. Also what doesnt help is the tyre model. But even with the current model, its a usable vehicle when driven properly. Fixing the tyre model and reducing steering lock with speed (as most video games do) are probably only dev changes required. User changes - ditch the keyboard for something with analogue steering, braking and acceleration. Oh. Hirochi Sunburst. That has traction control and stability control along with ABS. The D15 and grand marshal have ABS.
