Update 1.06 possible changes to the physics + general physics discussion.

[Imari]

This is a myth. The gyroscopic effect has very little to do with keeping a motorbike or bicycle upright. They proved this via installing counter gyroscopes and seeing how easy it was for people to ride. It made zero difference. The faster you go on a bike the easier it is to stay upright; this is because the effect of your steering adjustments is multiplied with speed. There is nothing more to it.

Do you have a source for this? I'd be interested to see.

A lot of a two wheeled vehicle's natural stability comes from the rake on the front forks, although there's stuff like penny farthings where obviously it's gyroscopic balance or nothing. Depends a lot on design, but modern designs are all mostly similar in that they are designed such that when they're rolling at speed if they tip to one side they innately want to stand themselves up again, just because of the geometry of the wheels.

At low speed however, a lot of that goes away or becomes negligible. In that case, I wouldn't be surprised if there's a difference between engine on and engine off on a motorbike, even if it's slight. I feel like I've had similar experiences, but it's been a long time since I rode a motorcycle.

 

[Mt. Lynx]

Except that reactive centrifugal effect doesn't act on the car. It's the force the car, or the tyres in this case, exert on the road/earth.



You have no idea what you're talking about. Go read Wikipedia, then come back and join us. I'm sorry, but you're not even on the same page.

The only way you could slot the words "centrifugal force" into that sentence is if you're talking about reactionary centrifugal force. If there's no reactionary centrifugal force, there's no centripetal force either and YOU CAN'T TURN.

The centrifugal force that injures people when you have an accident, presumably by impalement on the steering column or something similarly gruesome, is not a force. It's a convenient mental shortcut, but there is no force there at all. None.

The car stops (because the tree won't let it keep going straight).
The occupants, being not solidly attached to the car keep going at their previous speed until they encounter something solid enough to stop them, like the steering column.

The steering column definitely produces a force against your ribs, but not a centrifugal one. It's an imaginary force designed to make rotating frames of reference work, and it's completely unnecessary if you're considering the situation correctly.



Certainly. And what's producing the force that acts on that momentum? The tyre. Increased inertia means increased tyre load (to produce increased force) to get through the same turn.

Hmmmm...
Call it inertia then, I think they are the same. This could be a lingo thing too, english is not my native tounge, and I have never even studied "physics one'o'one" in english.
But either way, I thought it was (or that is how I interprated it) the opposing centripetal force that affected the tyres first, not the surrounding contact patch. In theory at least, I should be right, since tyres are flexible. In GT6 this is an entirely different matter though, since tyres seem not to flex.
And this is what I'm getting at. What if the physics model has cornering nailed completely, including the suspension? Without any flex in the tyres, the whole model becomes a mess - wich is the way it seems to be right now. Camber would be void (and so it is), the weight transfer under cornering load would be iffy, especially with rear heavy cars, since the back would have a much smaller operational window (this has been somewhat corrected, but I have no idea how).

So beside the terminology of the physics, could it be a reasonable possibility that our issues with the in game driving physics can largely be blamed on the absence of tyre flex?

 

[BrandonW77]

At first I felt no difference but I actually think input lag has been reduced, making the physics feel more intuitive. It may also be the physics that have been tweaked but I feel more control with the DS3 now.

The more I play the more I think this may be what we're noticing. Everything feels smoother and more direct, even just making small corrections on a straight where a change in physics wouldn't really be noticeable. A reduction in input lag would result in slightly more precise inputs which would make the cars feel slightly more responsive and well behaved.

 

[Johnnypenso]

You are not even reading what I write are you?

And no, once a car is in the air drivers have basically no control. Rally drivers need to setup their car before the jump, using the brakes to throw the weight on the nose if they want the nose of the car to dive. And GT does model the wind, even if PD didn't tell us about how they were doing it (basically you are saying they lied about the game... why?) you should know just by playing the game and watching the cars.

I can't explain it better than this:

Nope. Inertia affects tyre load. A heavier car requires more centripetal force to accelerate it around the turn.

Centrifugal force (as you're describing it) does not exist. It's a mental shortcut, and it's not appropriate when you're discussing physics seriously.

True centrifugal force is the force the tyres exert on the Earth.



Nope.

The car landed at the correct point (more or less) on the other ramp, and so was going at the correct speed. If he had gone faster he would have overshot, or undershot if he'd gone slower.

You can hear him cut the throttle in the video, and see the car pitch when he does. Why? Newton's Third: For every action there is an equal and opposite reaction.

Consider the car and wheel as two separate entities. They're connected, but are free to rotate with respect to one another. Directly after the jump they're not exerting any forces on each other, the wheel is spinning at whatever speed, and the engine/car is matching that speed.

When the driver lets off the throttle, you now have a force acting to decelerate the rotating wheel. Say we're looking at the left rear wheel. The deceleration provides a force on the wheel in a clockwise direction. And so there's an equal and opposite force acting on the car, twisting it counterclockwise around the rear axle.

Were the car on the ground, it would not be free to rotate and would simply slow down. But in the air there's nothing to stop the car rotating. Energy is not destroyed, always conserved. The energy that the rotating wheel had when it left the jump has simply been transferred into rotating the body of the car instead. The car is much heavier than the wheel, and so will rotate much slower, giving decent fine control of the attitude of the car.

This is why jumping cars is very tricky. It's not just about driving at X speed, because any moron could do that. It's about getting the speed right AND having perfect timing and precision to control the attitude of the car before landing.

Overall point being, in GT5/6 you can jump several hundred feet and land perfectly every time, no matter what you do with throttle and brakes on the ramp or in the air. This is not how reality works. In GT5 we could jump hundreds, maybe a thousand feet and still land on all 4 tires and keep on driving. Not how reality works. Gravity when you are in the air is not modelled accurately, the game engine ensures you land upright.

 

[Imari]

So beside the terminology of the physics, could it be a reasonable possibility that our issues with the in game driving physics can largely be blamed on the absence of tyre flex?

Not really. It's only in the last five or so years that I remember hearing people starting to talk seriously about having flex in tyre models, and there have been good sims around for a lot longer than that. Camber should work with or without flexible tyres, it's largely the body roll of the car that it's countering.

So assuming body roll doesn't work correctly (which it obviously doesn't if camber doesn't work), you could still make camber behave somewhat normally simply by introducing a fudge factor that put maximum grip at a few degrees of camber instead of zero. It'd be completely arbitrary and not at all related to the car and driving that was going on as it should be, but at least the numbers would be more or less like real life.

Besides, every tyre model has "flexible" tyres to some degree. Otherwise you'd have two options: full contact when the tyre surface is absolutely parallel to the road, and basically no content the rest of the time because it's riding on an edge, assuming the tyre is modelled as a box or something similar. There has to be some calculation that is saying "when the tyre is within 3 degrees of the road surface you have 85% grip" and so on. That's the "squish" mode, for want of a better term, the way the tyre flattens onto the road under weight even when it's at a bit of an angle.

What has been the recent thing with flexible tyres in sims has been adding lateral flex, which adds a little additional lag between turning the steering and having the tyre bite, the ability for the tyre to roll onto the sidewall, and a few other bits and pieces. It's important for feel, and it does give a more natural driving experience, but you can have a reasonably competent sim without it.


The truth is, because GT doesn't show us telemetry we have nearly no idea what's going on behind the scenes. The few things that we have established, like the camber "bug" seem to indicate that the physical simulation isn't exactly conforming to reality, nor are PD testing these things to make sure that they work. Camber is a fairly fundamental part of a driving simulator, and having it non-functional raises all sorts of questions about the rest of the simulation aspects.

Are they simulated, or is the whole thing just a well-tuned fudge that happens to produce a convincing driving experience? Are we in the Matrix or in reality?


P.S. Well done for attempting this in a second language.

 

[eclipsee]

You have no idea what you're talking about. Go read Wikipedia, then come back and join us. I'm sorry, but you're not even on the same page.

Eh! you're not very polite. 👎

I know what I'm talking about, but you don't understand me or you don't want to.

Centrifugal force is the result of the inertia force when a centripetal force is applied to an object in movement, if there wouldn't be a centrifugal force when in motion, this would lead us to a fictional world without inertia, that's why I posted those weird statements.

Have a nice day!

 

[Mt. Lynx]

Not really. It's only in the last five or so years that I remember hearing people starting to talk seriously about having flex in tyre models, and there have been good sims around for a lot longer than that. Camber should work with or without flexible tyres, it's largely the body roll of the car that it's countering.

So assuming body roll doesn't work correctly (which it obviously doesn't if camber doesn't work), you could still make camber behave somewhat normally simply by introducing a fudge factor that put maximum grip at a few degrees of camber instead of zero. It'd be completely arbitrary and not at all related to the car and driving that was going on as it should be, but at least the numbers would be more or less like real life.

Besides, every tyre model has "flexible" tyres to some degree. Otherwise you'd have two options: full contact when the tyre surface is absolutely parallel to the road, and basically no content the rest of the time because it's riding on an edge, assuming the tyre is modelled as a box or something similar. There has to be some calculation that is saying "when the tyre is within 3 degrees of the road surface you have 85% grip" and so on. That's the "squish" mode, for want of a better term, the way the tyre flattens onto the road under weight even when it's at a bit of an angle.

What has been the recent thing with flexible tyres in sims has been adding lateral flex, which adds a little additional lag between turning the steering and having the tyre bite, the ability for the tyre to roll onto the sidewall, and a few other bits and pieces. It's important for feel, and it does give a more natural driving experience, but you can have a reasonably competent sim without it.


The truth is, because GT doesn't show us telemetry we have nearly no idea what's going on behind the scenes. The few things that we have established, like the camber "bug" seem to indicate that the physical simulation isn't exactly conforming to reality, nor are PD testing these things to make sure that they work. Camber is a fairly fundamental part of a driving simulator, and having it non-functional raises all sorts of questions about the rest of the simulation aspects.

Are they simulated, or is the whole thing just a well-tuned fudge that happens to produce a convincing driving experience? Are we in the Matrix or in reality?


P.S. Well done for attempting this in a second language.

Sounds reasonable.
I do have a small "but" though;
What if PD really did do their homework properly, when working with KW and Yokohama? If so, suspension and tyres are working accordingly. So what is wrong then? Weight transfer? Or is it because PD has kept some old "fudge" coding that doesn't work with new tyre and suspension model?
Whatever it is, it's frustrating!

 

[mistersafeway]

Sounds reasonable.
I do have a small "but" though;
What if PD really did do their homework properly, when working with KW and Yokohama? If so, suspension and tyres are working accordingly. So what is wrong then? Weight transfer? Or is it because PD has kept some old "fudge" coding that doesn't work with new tyre and suspension model?
Whatever it is, it's frustrating!

That seems possible, considering the ride height glitch returned.

 

[Imari]

Eh! you're not very polite. 👎

I know what I'm talking about, but you don't understand me or you don't want to.

Centrifugal force is the result of the inertia force when a centripetal force is applied to an object in movement, if there wouldn't be a centrifugal force when in motion, this would lead us to a fictional world without inertia, that's why I posted those weird statements.

Have a nice day!

No, I'm not very nice because you do not understand that centrifugal force, as you describe it, is a fiction. There is no such thing as "inertia force", inertia is the resistance of an object to change it's momentum. The entire motion of the object is described by it's momentum, it's inertia and the centripetal force applied. No centrifugal force, unless you're talking about reactive centrifugal force which is negligible in the cases we're talking about, and no "inertia force".

Having no centrifugal force means nothing in particular, and it certainly doesn't result in a "world without inertia". If you choose your reference frame appropriately, there's simply no need for a fictional force.

From Wikipedia, where you should have done your homework:

Centrifugal force is the apparent force that draws a rotating body away from the center of rotation. It is caused by the inertia of the body as the body's path is continually redirected. In Newtonian mechanics, the term centrifugal force is used to refer to one of two distinct concepts: an inertial force (also called a "fictitious" force) observed in a non-inertial reference frame, and a reaction force corresponding to a centripetal force.

You are talking about the "inertial force", which is a fictitious force used to make non-inertia reference frames function. It's entirely imaginary. There is no actual force there.

I don't see why you would insist on working in a rotating reference frame. Is it somehow easier to treat the vehicle as stationary and have forces acting to move the entire world around it? Rotating reference frames only work when what you're interested in is entirely inside the reference frame. I don't think anyone is much interested here in the physics of how the driver is pushed against his seat, or how his drink spills over. The interesting interaction is between the car, which is "rotating", and the road, which is not. Which of these is easier to define as the reference frame, the car or the road?

If you'd like to explain why a rotating frame of reference is preferable, be my guest.

Why is a non-rotating frame preferable? Because it's a damn sight easier to calculate forces on a car of known mass and dimensions than it is to calculate forces on a road, attached to a planet. Strictly, said planet is also a rotating frame of reference but it is assumed that an contributary forces from rotation are negligible, and so it can be treated as an inertial frame.

Sounds reasonable.
I do have a small "but" though;
What if PD really did do their homework properly, when working with KW and Yokohama? If so, suspension and tyres are working accordingly. So what is wrong then? Weight transfer? Or is it because PD has kept some old "fudge" coding that doesn't work with new tyre and suspension model?
Whatever it is, it's frustrating!

I would say the suspension isn't working properly, because that's the part of the car that should be controlling the camber angle and how it varies. It's odd, because they're obviously calculating the attitude of the car graphically, so the information about the relative positioning of the car and road surface has to be there. It's simply not being used to modify the grip of the tyres.

It's possible that the tyres wouldn't work properly either if put into a system that did do camber properly, but it's impossible to know. Longitudinal weight transfer works in GT6, at least to some extent, so I doubt that they got that right but not lateral weight transfer. It should be the same code, really, there's no reason to separate out longitudinal and lateral.

Honestly, there's reasonable evidence that Gran Turismo's tyre model is pretty rudimentary. The fact that it was a simple grip multiplier in GT5 made all sorts of cars with tyres of staggered widths drive pretty poorly. They appear to have fixed that in GT6, but I doubt that they went further than simply scaling the grip by the width of the tyre.

Tyre pressures should be a basic part of any tyre model, because they're a simple modification of the "squish" factor that describes the contact patch. If you have sidewall flex, they're an important part of that too. That Gran Turismo doesn't have tyre pressure leads me to suspect that they don't have a "squish" factor describing the contact patch, and that if the tyres were actually presented at an angle to the tarmac they wouldn't work. Camber is probably just another grip multiplier, a negative one, rather than a physical modification of how the tyres contact the road and allowing the repercussions of that propagate through the physics system.

All speculation, and I could be wrong. PD made a lot of fuss about their collaboration with KW and Yokohama, but we're yet to see much out of it. The driving feels good, but there's very little that would actually demonstrate that the tyre or suspension models are very advanced. For all we know it's the driving equivalent of a really great dish of pasta. Delicious, but pretty basic.

 

[nasanu]

Do you have a source for this? I'd be interested to see.

Much of the research is published in Journals I can't find online, but here is an article referencing some of the work:
http://www.news.cornell.edu/stories/2011/04/researchers-explain-why-bicycles-balance-themselves

Oh and have a rethink about the penny farthing, do you really think such a slowly rotating wheel is going to stabilize the mass of a person sitting so high? Would be like an elephant leaning against an ant.

I can't explain it better than this:

Overall point being, in GT5/6 you can jump several hundred feet and land perfectly every time, no matter what you do with throttle and brakes on the ramp or in the air. This is not how reality works. In GT5 we could jump hundreds, maybe a thousand feet and still land on all 4 tires and keep on driving. Not how reality works. Gravity when you are in the air is not modelled accurately, the game engine ensures you land upright.

But this is just not true. You don't land perfectly flat, not in GT6. And what evidence at all do you have for gravity not being modeled properly in the air? It is absolutely absurd to think that you believe they have a different gravity model for when you jump. I think you are taking assists or game limits like the fact that code intervenes to stop rolling and applying the effect of that to a physics model.

Simple fact is that if gravity were to be incorrect then cars jumping at eiger would fall at an incorrect rate. Gravity is a very simple force, it just pull equally everywhere (more or less, lets not get too bogged down in detail here). Is the rate at which cars fall from the jump incorrect? NO. So gravity simply can not be wrong.

Either show that cars drop at an incorrect rate or give up your attempt to could the discussion with false statements.

Again, the two opposing forces are centripetal and inertial - centrifugal force, as it is being used here, does not exist.

It does exist, what matters is the inertial frame of reference. It is what centripetal force is called in certain observational conditions. Remember relativity? The frame of reference matters.

 

[Imari]

http://www.news.cornell.edu/stories/2011/04/researchers-explain-why-bicycles-balance-themselves

They're referencing the trail balancing the bicycle, which is perfectly true, and I also mentioned it.

What it doesn't mention, is that at slow speed the forces due to the stabilising effect of the trail are significantly weaker than the weight of the bicycle. This is why if you roll a bike by itself it will stay stable, but it will fall over before it stops. There's a critical speed below which it is no longer a stable system.

Below this critical speed, the gyroscopic effect helps. It particularly helps from the engine, where engine speed can be largely decoupled from ground speed with clever use of clutch and brakes.

It does exist, what matters is the inertial frame of reference. It is what centripetal force is called in certain observational conditions. Remember relativity? The frame of reference matters.

What matters is that the frame of reference be non-inertial. Centrifugal force doesn't exist in inertial reference frames. Since non-inertial reference frames have an acceleration as standard, they require fictitious forces in order to apply the same rules that apply to inertial frames. A fictitious force is by definition a non-existence force that is added in order to make the calculations work. It's a complicated bugger factor, is all.

Reactive centrifugal force exists. It is used in inertial reference frames and is the counterpart to centripetal force. It's not helpful, because nobody cares that the car exerted 2000N of force on the Earth. That's enough to move it about three fifths of a poofteenth. Bugger all, in layman's terms.

Centrifugal force in non-inertial reference frames is a mathematical construct, it has no physical counterpart. It's possible to do the calculations in a non-inertial frame without reference to a fictitious force, even, just use the acceleration that is by definition part of the non-inertial frame. It's just that people are well trained to grok inertial reference frames, and so transforming a non-inertial into an inertial frame with a fictitious constant force lets people intuitively see the results more easily.

 

[Johnnypenso]

But this is just not true. You don't land perfectly flat, not in GT6. And what evidence at all do you have for gravity not being modeled properly in the air? It is absolutely absurd to think that you believe they have a different gravity model for when you jump. I think you are taking assists or game limits like the fact that code intervenes to stop rolling and applying the effect of that to a physics model.

Simple fact is that if gravity were to be incorrect then cars jumping at eiger would fall at an incorrect rate. Gravity is a very simple force, it just pull equally everywhere (more or less, lets not get too bogged down in detail here). Is the rate at which cars fall from the jump incorrect? NO. So gravity simply can not be wrong.

Either show that cars drop at an incorrect rate or give up your attempt to could the discussion with false statements.

I have explained it already and provided a video to boot, but I'll give you a shorter simpler version and hopefully you'll understand it. When a car jumps a ramp, any ramp, the front of the car is airborne before the back right? So the front of the car starts falling before the back right? Objects tend to fall at the rate of gravity, 32 Ft/s/s. A car with a 10 foot wheelbase going 200km/h over the big jump at Cape Ring for example, takes about .06 seconds from when the front tires leave the ramp until the back tires leave the ramp. Still with me?

When those front tires leave the ramp they begin to fall at 32ft/s/s. In .06 seconds that's just under 2 feet/s. So in the absence of any other forces in 3 seconds the front of that car will be 6 feet lower than the back. In GT5 there were jump tracks where you'd be airborne for 5 or more seconds meaning that the front of the car should fall about 10 feet, essentially flipping over every time onto it's nose.

Of course that's not what happens. In each case, you can still land on on all 4 wheels. This is why I believe that when you are airborne in GT, either gravity is different than when you are on all 4 wheels, or it's simply suspended and some type of dumbed down physics model takes over so we all don't crash every time we jump a car. The fact that you can jump 100 feet, 300 feet or 800 feet and land exactly the same way, is undisputable proof that gravity in GT6 when airborne is not modeled correctly or as I said, it's suspended in favour of consistent gameplay.

 

[Imari]

The fact that you can jump 100 feet, 300 feet or 800 feet and land exactly the same way, is undisputable proof that gravity in GT6 when airborne is not modeled correctly or as I said, it's suspended in favour of consistent gameplay.

It's proof that cars in Gran Turismo are still based on the sliding brick model, more than anything. Gravity behaves in this way if everything is calculated from a central point, and the car behaviours are merely modifiers of how that brick slides.

If the car is an independent object with inertia that is supported by four wheels each calculated separately, you might expect to see something approaching reality. The fact that we don't tends to be another piece of circumstantial evidence that the physics system in Gran Turismo is little more than an elaborate fudge job.

 

[mistersafeway]

Back on the original topic, ride height seems to work properly now (tested with BMW Z4 GT3 and Honda HSV Base Model). Is this news to anyone else or did I just miss something previously? It's just that if it's new, it could indicate a centre of gravity fix which would explain less rolling over and smoother handling from cars that people have reported. Maybe. Possibly.

 

[nasanu]

I have explained it already and provided a video to boot, but I'll give you a shorter simpler version and hopefully you'll understand it. When a car jumps a ramp, any ramp, the front of the car is airborne before the back right? So the front of the car starts falling before the back right? Objects tend to fall at the rate of gravity, 32 Ft/s/s. A car with a 10 foot wheelbase going 200km/h over the big jump at Cape Ring for example, takes about .06 seconds from when the front tires leave the ramp until the back tires leave the ramp. Still with me?

When those front tires leave the ramp they begin to fall at 32ft/s/s. In .06 seconds that's just under 2 feet/s. So in the absence of any other forces in 3 seconds the front of that car will be 6 feet lower than the back. In GT5 there were jump tracks where you'd be airborne for 5 or more seconds meaning that the front of the car should fall about 10 feet, essentially flipping over every time onto it's nose.

Of course that's not what happens. In each case, you can still land on on all 4 wheels. This is why I believe that when you are airborne in GT, either gravity is different than when you are on all 4 wheels, or it's simply suspended and some type of dumbed down physics model takes over so we all don't crash every time we jump a car. The fact that you can jump 100 feet, 300 feet or 800 feet and land exactly the same way, is undisputable proof that gravity in GT6 when airborne is not modeled correctly or as I said, it's suspended in favour of consistent gameplay.

You write so much yet say nothing about GT6. Want me to show you a picture of a car in GT6 landing on its front wheels first (and not just that but slightly tilting to the left)? You just keep writing and writing about something that is not true. Why is this?

Also, your simplistic view on jumping physics excludes reality (think just a little please... until the car reaches the top of its parabola it does not matter if the road is beneath it or not).

These have to be faked because all of these cars landed rear wheels first or flat:

How many more examples should I show? The childish 'gravity acts on the front first so cars can't land flat' view is massively flawed and does not reflect reality.


^^
Just watch... Cars landing flat, on their back wheels, on their front... Its the jump angle, speed, car balance etc that matters, not one simplistic view of physics.

They're referencing the trail balancing the bicycle, which is perfectly true, and I also mentioned it.

What it doesn't mention, is that at slow speed the forces due to the stabilising effect of the trail are significantly weaker than the weight of the bicycle. This is why if you roll a bike by itself it will stay stable, but it will fall over before it stops. There's a critical speed below which it is no longer a stable system.

Below this critical speed, the gyroscopic effect helps. It particularly helps from the engine, where engine speed can be largely decoupled from ground speed with clever use of clutch and brakes.



What matters is that the frame of reference be non-inertial. Centrifugal force doesn't exist in inertial reference frames. Since non-inertial reference frames have an acceleration as standard, they require fictitious forces in order to apply the same rules that apply to inertial frames. A fictitious force is by definition a non-existence force that is added in order to make the calculations work. It's a complicated bugger factor, is all.

Reactive centrifugal force exists. It is used in inertial reference frames and is the counterpart to centripetal force. It's not helpful, because nobody cares that the car exerted 2000N of force on the Earth. That's enough to move it about three fifths of a poofteenth. Bugger all, in layman's terms.

Centrifugal force in non-inertial reference frames is a mathematical construct, it has no physical counterpart. It's possible to do the calculations in a non-inertial frame without reference to a fictitious force, even, just use the acceleration that is by definition part of the non-inertial frame. It's just that people are well trained to grok inertial reference frames, and so transforming a non-inertial into an inertial frame with a fictitious constant force lets people intuitively see the results more easily.

You are just starting to argue semantics. It is not constructive.

As for the bike, did you read the article? It was specifically debunking what you claim it supports. Besides, how much gyroscopic force do you really think is being generated at < 5mph? Pretty much none.

 

[GTP_CargoRatt]

Back on the original topic, ride height seems to work properly now (tested with BMW Z4 GT3 and Honda HSV Base Model). Is this news to anyone else or did I just miss something previously? It's just that if it's new, it could indicate a centre of gravity fix which would explain less rolling over and smoother handling from cars that people have reported. Maybe. Possibly.

RH is still reversed, on most cars anyway, just like it was before the update. Nothing has changed as far as that goes, that I have seen anyway. Still setting my cars up with higher front/low rear for better rotation.

 

[Johnnypenso]

You write so much yet say nothing about GT6. Want me to show you a picture of a car in GT6 landing on its front wheels first (and not just that but slightly tilting to the left)? You just keep writing and writing about something that is not true. Why is this?

Also, your simplistic view on jumping physics excludes reality (think just a little please... until the car reaches the top of its parabola it does not matter if the road is beneath it or not).

These cars have other forces acting on them that don't exist in GT6. Did you not notice the "absence of other forces" in my response? As others have noted, the spnning motor and rear wheels help to keep a car's nose pointed up. This does not happen in GT6. You can wind the motor right out, or you can slam on the brakes and stop everything from spinning, and nothing changes, which means those forces are not modeled.

​

 

[Imari]

You write so much yet say nothing about GT6. Want me to show you a picture of a car in GT6 landing on its front wheels first (and not just that but slightly tilting to the left)? You just keep writing and writing about something that is not true. Why is this?

Also, your simplistic view on jumping physics excludes reality (think just a little please... until the car reaches the top of its parabola it does not matter if the road is beneath it or not).

These have to be faked because all of these cars landed rear wheels first or flat:

How many more examples should I show? The childish 'gravity acts on the front first so cars can't land flat' view is massively flawed and does not reflect reality.

Back to the jumping thing, in real life you can make a car land pretty much however you want to, provided you've got the hang time to adjust the car. They can land flat, nose down, tail down, roof down or whatever else you like. It might be tough to get roll on them without a canted takeoff, but otherwise it's all down to how the driver wants to control the car in the air.

In Gran Turismo once you take off, you're a passenger until your wheels are on the ground again. How you land is determined at the point of takeoff, and that's not how real life works.

 

[FussyFez]

I like this thread, I'm learning stuff.

My 2 cents...

On my 360@eiger jump, I was full throttle leaving the jump, the entire time I was rotating, and on landing. I landed pretty flat, and the full throttle landing made sure the rear kept sliding on landing.


I'm pretty sure, if this happened in real life, as soon as I left the ramp, I would have to hit and hold the clutch, keep the revs up, then dump the clutch in conjunction with full throttle a split second before landing. This (in my head atleast) should keep the car flat for the rotation in the air, but then provide me with the rear wheel speed to keep the drift on landing.


If I used my current in game technique in real life, the car would land on its back wheels (or bumper) and I'd likely go spearing off in a random direction.



Also, I'm no physics expert, but I would agree that centrifugal force is simply a layman's term for inertia, and how it's perceived when the subject is traveling in an arc.

 

[nasanu]

Back to the jumping thing, in real life you can make a car land pretty much however you want to, provided you've got the hang time to adjust the car. They can land flat, nose down, tail down, roof down or whatever else you like. It might be tough to get roll on them without a canted takeoff, but otherwise it's all down to how the driver wants to control the car in the air.

In Gran Turismo once you take off, you're a passenger until your wheels are on the ground again. How you land is determined at the point of takeoff, and that's not how real life works.

Can you show me a single example of a car changing pitch on its own in the air? Because I don't believe it in the slightest and I have never seen it. Rally drivers certainly cant do it, once they take off wrong its all over (or perhaps those on here are just far better drivers?). Just one single video, only then I'll believe a few spinning wheels or spinning components of an engine can overcome the entire mass and inertia of the car to change its path in a split second.

 

[Dodzzz]

I'm with nasanu

To make a successful long jump, you need a perfectly calculated everything, car spring rate, weight balance, speed when airborne, length and angle of the ramp, wind speed, and etc. What shows in the world record fail video posted earlier, is not that he lift off the throttle that caused the car to plunges down front first, it was just a coincidence that the car is actually start plunging down after it is losing the spring force generated earlier when the car jumps right after you hear the engine is cut.

It was just a FAIL attempt at trying to break world record, not just because a mere fault of the driver lift off his shoe.. but it's all about the whole calculations that went all wrong.

Edit: I found this:



If you listen and watch the video above carefully, the car nose angle doesn't change a bit even after he releases the throttle.. and landed perfectly.. it was just perfect calculation imho.

 

[MatskiMonk]

I love Physics.

I was going to weigh in on the discussion, but I couldn't be massed. So I did this instead.

Although, engine rotation must affect cars pitch/roll depending on engine orientation, or else why would cars wobble/lean when you rev them at a standstill?

Also, real word =/= GT, so perhaps we try coming up with more ways to test it. Like allowing the car to roll down a hill in 1.06 vs. previous version, and just observe if their is a difference thanks to gravity?

 

[Griffith500]

There is no difference in gravity, nobody with any sanity has been arguing that.

What the argument has become is that GT doesn't model free-body physics correctly, and only gives a passable representation of real-world physics if all four wheels are on the ground.

Obviously, the simple model it uses for body contact, for example, means that the car can drop at the correct 9.81 m/s² because that's a constant force and even the most basic "physics model" can accommodate that.

It takes anything but zero force to "overcome" inertia (friction is a different matter), and suddenly stopping 20 kgs of wheel from rotating at 60 mph will exert a force on the car that will cause it to rotate at an appreciable rate, if it is free to do so - otherwise that rotation becomes a force imparted onto the road (via the chassis, suspension and wheels) instead. It's a well known trick amongst dirt jumpers, and it's even saved me from looping out on occasion.


But that isn't really the crux of the issue, it's simply that the forces are interpreted according to their influence on the tyre contact patches only, and not on the actual physical objects themselves, which would then be communicated to the contact patches via the interactions determined by the couplings in the physical model.

As an example, you can put your car on its roof / roll-hoop in games like GPL and NR2003, with the wheels still spinning, brush the brakes and watch the resulting torque rock the car gently. That's a natural result of just modeling things properly (and somewhat expensively for the hardware at the time, in GPL's case). That can't happen in GT, because the wheels can't push against fresh air.

 

[FussyFez]

@nasanu

I know it's RC but it's the same physics at work.

 

[Johnnypenso]

If one doesn't understand physics that's ok. But to simply ignore the physics behind this because you don't understand it, don't like it or don't believe it, doesn't make for much of a discussion

 

[Imari]

@nasanu

I know it's RC but it's the same physics at work.

That's a better demonstration than I would have thought it possible to find, well done!

 

[nasanu]

@nasanu

I know it's RC but it's the same physics at work.

Are you serious? You could not possibly find a worse way to try to prove your point. An RC trucks weight is almost entirely wheel and engine mass. Of course that has an impact. On a real car we can discount the rotational mass of the engine because we can see it rotates the wrong axis, just sit and rev an car engine, it tilts the car side to side not front to back. So we are left with the wheel mass as the active force when using the cars controls to alter it in the air. In the RC truck that wheel mass is likely to be over 50% of the total mass. What would the wheel mass of a road car be? Even 5%? I doubt even that.

As I said, find me one singe example of this happening in a real car.

 

[shirakawaa]

If you still don't believe it has an effect, look at what happens to a car's suspension during a stationary dyno run when it accelerates/decelerates. Example:

 

[Griffith500]

Are you serious? You could not possibly find a worse way to try to prove your point. An RC trucks weight is almost entirely wheel and engine mass. Of course that has an impact. On a real car we can discount the rotational mass of the engine because we can see it rotates the wrong axis, just sit and rev an car engine, it tilts the car side to side not front to back. So we are left with the wheel mass as the active force when using the cars controls to alter it in the air. In the RC truck that wheel mass is likely to be over 50% of the total mass. What would the wheel mass of a road car be? Even 5%? I doubt even that.

As I said, find me one singe example of this happening in a real car.

You'll have to run the numbers for us, I think. Don't forget that the rotary force from the wheels stopping or speeding up is proportional to both the mass of the wheel and its acceleration, and is sustained throughout that acceleration. Then you'll need to show us just how much force is required to rotate a car in mid air at the required rate.

I outweigh my bicycle by a factor of 4, the wheels constitute less than a fifth of the bike's weight. So braking the wheels from about 20 mph is enough to rotate a mass 20 times their size relatively quickly enough to turn a looping-out situation into a nose-dive; just as a ball-park starting point. The trick works with heavier riders on lighter bikes, too, implying a mass ratio greater than 20, probably greater than 30, is still useful in such situations. These car jumps we're witnessing last much longer and involve much smaller angles to the horizontal than a typical bicycle jump, so even higher ratios (lower resultant rotation rates) can be considered useful; but there's far more acceleration to play with on the braking aspect because of the higher speed, too, by a factor of at least 2, maybe 3. The mass ratio on an RC car is closer to 4 or 5, not less than 1 as you imply, with tight constraints on flight time and angles, as with a bicycle.
What's the mass ratio for a car?



Never mind that the video Johnnypenso posted, and the one that Dodzzz posted supposedly refuting the idea, both clearly demonstrate "aftertouch". You can even see how the trophy truck has rolled slightly due to having throttled on after having braked to bring the nose down. That engine torque roll effect, as you say, even works when the car is supported by the road and opposed by the suspension, never mind in the open air; plus it can only be proportional to the acceleration felt by the rotational mass of the engine (and that due to any friction), the same goes for any connected drivetrain parts.

The rotational mass of the engine and flywheel is probably comparable to that of a couple of wheels. If the car is in gear, any throttle blips will speed up the wheels as well, and they do rotate in the "correct" plane; that rotation will be opposed by the chassis through the suspension mounts. That means the chassis will rotate in free air and the "opposition" will therefore be a purely inertial force, the result of accelerating the chassis in the direction of wheel rotation. Neglecting air resistance effects.



Now, this after-touch is of limited use in cars, for obvious reasons (no-one really gets that much air...), what's generally more important on rotation rates in mid-air (and helps to avoid the need for after-touch) is controlling the difference in contact patch forces around take-off, which you can influence with brakes or throttle via "weight transfer", but watch out for suspension interactions...

The point, really, is that a full simulation should be able to reproduce these effects via inertial interactions, but that is not how the physics engine appears to be built; that construction has other, more profound effects than simply not being able to rotate a flying car, because it affects any forces being put into or "generated" by the car. Those effects only seem to be accounted for via the contact patches, hence the weird behaviour the moment any of them aren't touching something.

 

[FussyFez]

Ignorance is bliss, evidently.

I have a 1:10 touring car, the wheels and tyres make up less than 5% of its mass.

It behaves in exactly the same way as the RC truck video.

The effects are probably more pronounced in the RC stuff (relative torque to mass etc, I have an RC motor with 119nm or torque) , but again, it's all the same physics at work.


I'm not here to convince you.

If you don't believe me, I don't care really.

Others here have a much better understanding and explanation of why it happens, than I could ever give.