Physics "bug" or the real thing?

[amfitamine]

When coming to a full stop at the top of a steep corner like in Daytona or the "High Speed ring", parrallel to the track, I noticed my car (well different cars) to slide sideways, as if on ice, down to the lowest part of the track. I use the Professional mode with all assists off. I put my car into nutral (G25 + clutch) but using the PS3 controller and applying the brake will do the trick just the same. Now I have never been to Daytona or on any other track (yet) but I do drive a car for about twenty years now and this just doesn't feel wright..as if the low speed or no speed simulation is way off. What do you think?

BTW changing the simulation mode doesn't help (tires are S2)
(Once, at the top of the corner/track, I let the car, in nutral, rock back and forth for a while just by keeping the steeringwheel at the same angle. There doesn't seem to be any friction....)

Excuse my English....

 

[H3rmaN]

most likely a bug! Unless the angle of the bend is over 45 degrees I highly doubt that would happen in real life! Even with normal road tyres, nevermind with sports tyres!

 

[750F-RM]

i've seen this happen on steep banked add-on tracks on GTR2 (great pc sim). surely a bug, no way that'd happen irl

 

[Nachous]

Any vid about this in youtube?

 

[deadlyz33]

Sounds like a bug to me. Hopefully it will be fixed.

 

[FoolKiller]

Daytona has a 31 degree bank and if I recall there is a minimum speed you have to keep up in order to "stick" on the banking.

I believe it is the Days of Thunder ride at King's Island amusement park where they give facts about NASCAR and Daytona with Richard Petty and they mention it. I believe it was something like 60mph.

 

[amfitamine]

I'll see if I can post a VID on Youtube tomorrow!

 

[deadlyz33]

Daytona has a 31 degree bank and if I recall there is a minimum speed you have to keep up in order to "stick" on the banking.

I believe it is the Days of Thunder ride at King's Island amusement park where they give facts about NASCAR and Daytona with Richard Petty and they mention it. I believe it was something like 60mph.

Don't believe all you see on tv. You think a car will slide on a 31 degree bank, not to mention if its sitting on slicks. Ever regular tires. Haven't you ever seen San Fransisco? There streets are atleast 30 degrees and you don't see parked cars sliding to the bottom.

 

[red7]

This has been going on since gt4 it might be slightly better now but still, with all the attention pd pays on other details, like car modeling, this is abit unacceptable IMO.

Also try spinning your car or just sliding to a stop and watch your tires just before your car comes to a stop. GT tire physics are alittle curious at times.

 

[cjmodiano]

I love the people who try to justify everything they see in gran turismo...

Anyway, I wouldn't call this a bug, but just a weakness of the physics engine.

 

[deadlyz33]

I love the people who try to justify everything they see in gran turismo...

Anyway, I wouldn't call this a bug, but just a weakness of the physics engine.

Yup, Maybe instead of having realtime physics they simply adjust traction ratings of each tire per car depending on a bunch of variables.

 

[danjama]

i will probably be the only thinking this but i am sorry, why should PD have to think about people who want to sit still on a 31degree bank in a racing game? It just isn't their problem to worry about that. Fair enough the tyres may act differently to variables, but why the hell are u sitting still in a racing game? They have done bloody well with GT5P and i dont see why something as small as that is important. Just drive as fast as you can around it and you wont notice!

 

[amfitamine]

Danjama,
Isn't it possible that the same physics engine flaw making a car slide going 0 mph could affect the cars handling at 100 mph? Besides, PD is calling GT5 a simulation and not a "simulation only if you drive 100mph or faster". Noticing these "small" issues and naming them gives PD the change to further enhance their product. Love the game and I am sure PD will tune the engine to a point where it can trully be called a simulation.

 

[Flerbizky]

This has been going on since gt4 it might be slightly better now but still, with all the attention pd pays on other details, like car modeling, this is abit unacceptable IMO.

Almost as unacceptable as actually stopping on the banking at Daytona.

 

[gitSehT]

Don't believe all you see on tv. You think a car will slide on a 31 degree bank, not to mention if its sitting on slicks. Ever regular tires. Haven't you ever seen San Fransisco? There streets are atleast 30 degrees and you don't see parked cars sliding to the bottom.

There is a huge difference between a car parked facing up or down a hill, compared to sides ways. The contact patch is greatly reduced! I am not saying the physics are correct but, I would not want to try to park my car across a hill in san fran.

 

[TheCracker]

There is a huge difference between a car parked facing up or down a hill, compared to sides ways. The contact patch is greatly reduced! I am not saying the physics are correct but, I would not want to try to park my car across a hill in san fran.

The contact patch is exactly the same, no matter which way the car is facing.

There's no way in real life that a car would just slide sideways down an incline as shallow as the Daytona banking.

 

[Flerbizky]

The contact patch is exactly the same, no matter which way the car is facing.

There's no way in real life that a car would just slide sideways down an incline as shallow as the Daytona banking.

You can call it a lot of things, but shallow isn't exactly fitting...

A slice of the 31 degree banking :

 

[TheCracker]

IMO It's too shallow for the weight of the car to overcome the stationary lateral grip of the tyres.

 

[fasternfaster]

if you roll the car back and fourth on a banked incline because there is a SIDEWAYS force on the contact patch a certain SLIP ANGLE develops. The car will move in the direction of the average slip angle. Turning the steering wheels would affect the slip angle though.

If the car is indeed ROLLING then this is no bug, but real physics. if the car is standing still and still sliding, then yes this is a bug. FYI: i dont have the game.

 

[amfitamine]

The strange thing is when you face the car up or down the incline and apply brakes it doesn't move it only slides sideways(with brakes)when parallel to the track. When under an slight angle (down) the front tires will slide down while the back stays in place. Makes me think the engine has a problem with sideways grip/friction at (low or) zero speed or maybe this is what makes sliding work at higher speeds.....

 

[Zer0]

The contact patch is exactly the same, no matter which way the car is facing.

Not the same:

The slide in GT5 seems excessive but who knows.. ?

 

[fasternfaster]

you are crazy.. thats not what a contact patch is. thats the shape of the skid mark the tires leave..

a contact patch on a properly inflated tire is ALWAYS an oval. if its anything else then you tires are not round.

its true if you move the oval one way(sideways) vs straight it will leave a thinner mark on the ground. But at any moment in time the same surface area is making contact with the ground. thats why its called a contact "patch"

 

[amfitamine]

The "funny" part of the "rolling back and fourth" thing I mentioned is that a car looses energie through friction and should come to a stand stil within a few "rolls". In GT there didn't (doesn't) seem to be any loss of energy, it just kept rolling back and fourth. Looked at it for maybe 6 rolls (all the same in hight) when i stoped the test. I'll see if I can reproduce it again.

 

[fasternfaster]

could it have been a pendulum type of roll?

 

[amfitamine]

"could it have been a pendulum type of roll?" That is exactly what it was. But that should stop right? Else you would havesome kind of perpetuem mobile.

 

[fasternfaster]

it wouldnt stop untill it gets to the bottom of the hill. physics sound legit.

As long as holding the brakes down stops the car, there is no bug.

 

[amfitamine]

it doens't pendle to the bottom, It just stays at the same hight. Pandeling back and fourth. When you do apply the brake, at parallel position to the track, it slides down.

 

[gitSehT]

Tires are made to preform in a straight line, park your car on angle (in real life) and watch the side wall compress. If the weight of your tire is entirely on the one side of your tire your contact patch is greatly reduced. This is also one of the reasons why you should brake in a straight line, when your tires are turned the contact patch is reduced.

Another problem is weight transfer, the tires on the low side of the car have more weight to support then the tires on the high side.

 

[fasternfaster]

it doens't pendle to the bottom, It just stays at the same hight. Pandeling back and fourth. When you do apply the brake, at parallel position to the track, it slides down.

Based on that description yes it is a bug. In fact there are 2 bugs.

One is that there is no rolling resistance in the tires.
Two that cars slide sideways when brakes are applied.

 

[cleanme]

I can explain it to you all. It lies in the mathematical equations that are used to "describe" car physics. One of the best formulas to use for grip and an object with 4 contact points is called the "Pietzchke Formula."

This formula basically determines how much grip there should be. It works amazing at speed, the faster you go the more accurate! Now you're saying, wait usually as things increase the less predictable/accurate. With the Pietzchke Formula the inverse is true.

In the Pietzchke formula speed or velocity is in the denominator of the formula. Big Problem!! As the car's velocity slows, the velocity becomes closer and closer to ZERO. What happens when you divide by ZERO? Error, error! Undefined, impossible!

Sooooo, to get over that they (car game makers) use a second formula to simulate grip&contact for lower speeds. In older games you could actually feel where your car was being controlled by one formula and then switching over to the other. This is the case here!

It is extremely hard to simulate grip when v=0, hence the sliding problem. There you have it.