Why is GT7 treating all gearboxes like automatics?

[km666]

I'm not sure why you think it's an extreme overrev scenario. You claimed that your car loved to be oversped, but that's false. Overspeeding your engine does nothing good to your car in any way, not even if it's just slightly above the rev limit.

Centripetal force is proportional to the square of the angular velocity. Going from 7000 rpm to 8000 rpm increases the stresses by 30%. Material fatigue in steel is a function of stress and the number of cycles – the higher the stress, the fewer cycles it can handle.

Material failure is stochastic (can be described by random probability distribution). If the engineers picked 7000 rpm as the redline they knew they had to design the components so that they would have a certain safety margin against failure over the design life of that engine given the stresses at 7000 rpm and the number of cycles the engine was thought to have at that speed over its design life.

So when you drive your engine at a higher speed than it's designed for you are increasing the probability of failure and reducing the lifespan of the engine.

Driving the engine at high speeds below the rev limiter for extended periods of time will also reduce its lifespan and increase the probability of failure if you exceed the number of cycles that the engine was designed to have at those speeds. Redlining the engine for a few seconds is safe. Redlining for a minute might be ok if the engine and the oil is in good condition. Redlining the engine for an hour is not a good idea. Even if the engine survives that you might want to have it replaced afterwards.

So this idea that your car loved to be overrevved is ridiculous. Overrevving an engine does nothing good for the car, rather the opposite.

Glad to see you've come to accept the part about how it could damage your drivetrain.

Ay? As I said tuning the computer will raise the ceiling when revving up, which on an na would in simple terms be increasing an arbitrary rev limit and increasing fuelling to match injector requests and cooling requirements. But we're talking about decreasing revs where the cooling element of adapted feulling is irrelevant as its in its overrun cycle and bypassing the injectors feeding straight back down the line. If the hardware can handle an increased rev ceiling from software alone it can handle it full stop. Yes when you start tuning the suggested oil changes go out the window but I wouldn't want to buy a car off any fool who thinks a motor can do 20k between changes.

Engineers are bound to all sorts of constraints when bringing a car to market. The number any given manufacturer settles on is purely arbitrary. For instance from my 20v turbo vw days having the rev limit set much past 5'500rpm is pure marketing bollocks as the turbo has given its best long before then and you're losing power wringing it out.

That whole section about rev limiters looks very outdated. We're a long way past 8v ford engines running mineral oil pretending they can handle 10'000 mile oil changes. Any engine that cannot survive being redlined for one minute, is quite frankly an engine that should've been scrapped before it left the factory.

In an extreme scenario, like dumping it into 1st at motorway speeds is pretty much going to get you a load of neutrals.

 

[daan]

In an extreme scenario, like dumping it into 1st at motorway speeds is pretty much going to get you a load of neutrals.

Which is what GT used to allow and now it doesn't, so I'm not exactly sure what the argument is anymore.

 

[eran0004]

Whatever was built in by Suzuki. It was a bit of an aggressive downshift (some say near crash)

Remember, the front brake and throttle are on the same bar. I managed to pull the lever and pin the throttle for a moment.

I see. But in that scenario you weren't engine braking, instead your engine was actually producing power and the rev limiter kicked in to cut the power.

As I said tuning the computer will raise the ceiling when revving up, which on an na would in simple terms be increasing an arbitrary rev limit and increasing fuelling to match injector requests and cooling requirements. But we're talking about decreasing revs where the cooling element of adapted feulling is irrelevant as its in its overrun cycle and bypassing the injectors feeding straight back down the line.

We're talking about engine stresses and overspeeding - not about how to tune an engine to produce more power.

If the hardware can handle an increased rev ceiling from software alone it can handle it full stop.

The fact that you can raise the rev limit doesn't mean that the engine can handle the increased limit.

Engineers are bound to all sorts of constraints when bringing a car to market. The number any given manufacturer settles on is purely arbitrary.

It's not arbitrary at all. A high revving engine is subject to greater stresses than a low revving engine and as such requires stronger components to cope with those stresses. The engineers design the components based on what type of engine they are building and pick a suitable rev limit for it. You don't design a 6000 rpm engine the same way as an 8000 rpm engine, thay would be a waste of money.

For instance from my 20v turbo vw days having the rev limit set much past 5'500rpm is pure marketing bollocks as the turbo has given its best long before then and you're losing power wringing it out.

The rev limiter is there to stop you from destroying the engine. It's not there to tell you when to shift gears.

That whole section about rev limiters looks very outdated. We're a long way past 8v ford engines running mineral oil pretending they can handle 10'000 mile oil changes. Any engine that cannot survive being redlined for one minute, is quite frankly an engine that should've been scrapped before it left the factory.

It seems line you misunderstood that section. I wrote that redlining an engine for a minute will probably be fine (remember, material strength is stochastic, we're talking about the probability of failure) provided that the engine and the oil is in good condition. If you want to redline an engine in bad condition for a minute straight you're welcome to take your chances with that, but we can't say that it will probably be fine anymore.

In an extreme scenario, like dumping it into 1st at motorway speeds is pretty much going to get you a load of neutrals.

Okay? How is that relevant?

 

[shadow0460]

It's not. Shifts on the racing "manual" trans take longer to complete.

 

[Voodoovaj]

It's not. Shifts on the racing "manual" trans take longer to complete.

Correct. I've had instances where I've gone from 6th to 2nd, and it gets close, but it never "bounced" off the limiter. Touched, but never bounced.

 

[km666]

We're talking about engine stresses and overspeeding - not about how to tune an engine to produce more power.

The fact that you can raise the rev limit doesn't mean that the engine can handle the increased limit.

It's not arbitrary at all. A high revving engine is subject to greater stresses than a low revving engine and as such requires stronger components to cope with those stresses. The engineers design the components based on what type of engine they are building and pick a suitable rev limit for it. You don't design a 6000 rpm engine the same way as an 8000 rpm engine, thay would be a waste of money.

The rev limiter is there to stop you from destroying the engine. It's not there to tell you when to shift gears.

It seems line you misunderstood that section. I wrote that redlining an engine for a minute will probably be fine (remember, material strength is stochastic, we're talking about the probability of failure) provided that the engine and the oil is in good condition. If you want to redline an engine in bad condition for a minute straight you're welcome to take your chances with that, but we can't say that it will probably be fine anymore.

I dont think you understand engines, you seem to think that cars are bought to market operating at thier absolute limit and that's simply not the case at all. Every car has a different 'ceiling' and the whole idea that over engineering is a waste of money makes no sense at all when there are plenty of examples of cars being engineered;
To take a lifetime of abuse and missed maintenance (any V8 that produces less than 40bhp per litre) Future homologation purposes (2JZ, RB26, EJ20, 4g63)
Or just detuned like hell for marketing purposes or local emissions standards (M57 etc)

The rev limiter serves a different purpose in every application. You only have to look at power curves to understand that. My example of a Honda Vtec is still climbing where the limiter calls it a day. The 20v turbo actually can flow quite well at high revs but is choked by the tiny turbo running stupid amounts of boosts for the application, so is losing power by that point.

Oil technology is such that provided a car is run on fully synthetic it can still run, and survive, with residual oil only. I think it was Ford who found that some old Escort or whatever could manage over a thousand miles dry.

If Engineers worked as you thought the following should be impossible, there is no change to the hardware beside the deactivation of two ancillary emissions devices which have no effect on the tailpipe emissions. A 40-something percentage gain in outright power and almost exactly a 33% gain in torque. You might notice that the torque arrives earlier and plateaus more gently while the power curve means no need to hold onto revs, hence the reduced rev limit in this application. This is likely very near the absolute limit for the standard clutch, but it still hasn't been touched 15'000 miles later on.

 

[eran0004]

I dont think you understand engines

You're wrong.

you seem to think that cars are bought to market operating at thier absolute limit

I haven't said anything like that. I said that engines are designed according to intended use and design lifetime. Nobody builds an engine that's intended to be redlined excessively and absolutely not to be overrevved by shifting down too soon.

Every car has a different 'ceiling' and the whole idea that over engineering is a waste of money makes no sense at all when there are plenty of examples of cars being engineered;
To take a lifetime of abuse and missed maintenance (any V8 that produces less than 40bhp per litre) Future homologation purposes (2JZ, RB26, EJ20, 4g63)
Or just detuned like hell for marketing purposes or local emissions standards (M57 etc)

You don't think the engines were made for those scenarios? I also noticed the lack of an example where an engine has been designed to be overrevved without causing any additional wear or reducing it's lifetime.

The rev limiter serves a different purpose in every application.

No it doesn't. Its only purpose is to prevent overspeeding the engine.

You only have to look at power curves to understand that.

It has nothing to do with the power curve.

My example of a Honda Vtec is still climbing where the limiter calls it a day.
The 20v turbo actually can flow quite well at high revs but is choked by the tiny turbo running stupid amounts of boosts for the application, so is losing power by that point.

What's the relevance?

Oil technology is such that provided a car is run on fully synthetic it can still run, and survive, with residual oil only. I think it was Ford who found that some old Escort or whatever could manage over a thousand miles dry.

An engine that breaks after a thousand miles is hardly tolerable and it certainly doesn't prove that an engine can be safely overrevved without causing additional wear.

If Engineers worked as you thought the following should be impossible

How so?

 

[km666]

You're wrong.

An engine that breaks after a thousand miles is hardly tolerable and it certainly doesn't prove that an engine can be safely overrevved without causing additional wear.

These two statements contradict one another. Your 'engineering kowledge' seemed to be based in another century or steam power or something.

There is no such thing as excessive redlining, it is one part of the engines operating speeds. If an engine were unsafe to rev the redline would be set lower.

 

[eran0004]

These two statements contradict one another. Your 'engineering kowledge' seemed to be based in another century or steam power or something.

I don't see any contradiction, care to elaborate?

I have an engineering degree from 2022, but you're absolutely correct that a lot of our knowledge on how materials behave stem from research made in the 20th century.

There is no such thing as excessive redlining, it is one part of the engines operating speeds. If an engine were unsafe to rev the redline would be set lower.

Again, material fatigue is a function of stress and the number of cycles. When steel or aluminium is subject to high stress, microscopic cracks begin to form in the material. With each cycle, more cracks are formed until eventually the material fails. The higher the stress, the fewer cycles the material will last, see example below of a fatigue test of steel.

The engine's design life is based on a normal use case and excessive redlining will drastically shorten it. For example, your daily driver's engine might be rated for 100 000 km but if you drive that car hard at racetracks instead it might only last a few hundred hours before some engine component fails.

As for the redline and engine damage, you're correct that they are related to each other. Now you just need to come to terms with the fact that material science is stochastic (can be described by probability theory) and that materials can be subject to fatigue, so that increasing the stresses on the engine increases the probability of failure and reduces the expected lifetime.

 

[km666]

I don't see any contradiction, care to elaborate?

I have an engineering degree from 2022, but you're absolutely correct that a lot of our knowledge on how materials behave stem from research made in the 20th century.

Again, material fatigue is a function of stress and the number of cycles. When steel or aluminium is subject to high stress, microscopic cracks begin to form in the material. With each cycle, more cracks are formed until eventually the material fails. The higher the stress, the fewer cycles the material will last, see example below of a fatigue test of steel.

View attachment 1506222

The engine's design life is based on a normal use case and excessive redlining will drastically shorten it. For example, your daily driver's engine might be rated for 100 000 km but if you drive that car hard at racetracks instead it might only last a few hundred hours before some engine component fails.

As for the redline and engine damage, you're correct that they are related to each other. Now you just need to come to terms with the fact that material science is stochastic (can be described by probability theory) and that materials can be subject to fatigue, so that increasing the stresses on the engine increases the probability of failure and reduces the expected lifetime.

What kind of engineering degree? I'm staggered that you have one and believe that engineers have the final say in any aspect of the automotive industry. That software tweaks can unlock almost an extra 100bhp in a car starting with 240bhp just goes to show how engineering purity of concept gets overridden for the sake of marketing, model hierarchy or accountants.

 

[benjoi84]

It's a game

 

[Greycap]

There is no such thing as excessive redlining, it is one part of the engines operating speeds. If an engine were unsafe to rev the redline would be set lower.

Redlining, as in flooring it until the ECU says "no more", is mostly safe for short periods. However you're completely ignoring that @eran0004 is talking about overrevving, which is an entirely different thing. You can't overrev a healthy modern engine by throttle alone as the limiter will literally limit it but you absolutely can by downshifting when already at the redline.

That's what people have been doing in GT for ages and it can kill an engine, I know a guy who toasted his Fiat by doing just that. Accelerated hard and then missed the shift to fifth at the redline and threw in third instead, resulting in measurably stretched connecting rods and damaged valves as the springs couldn't close them fast enough to escape the extra 3000 rpm or so that the engine wasn't designed for. It could take the 7000ish rpm redline but not the around 10000 rpm the misshift caused.

To put that into perspective, he put in one gear too low and the engine was done for. People have been putting in two or three gears too low in GT, not only should it be physically impossible (as has been done in the game now) as no synchros allow doing it but a real engine would throw its inner workings on the road in an eyeblink or strip the gears of teeth, or both if it could be forced into gear like that.

 

[eran0004]

What kind of engineering degree?

Civil engineering.

I'm staggered that you have one and believe that engineers have the final say in any aspect of the automotive industry.

That's not something I have said.

That software tweaks can unlock almost an extra 100bhp in a car starting with 240bhp just goes to show how engineering purity of concept gets overridden for the sake of marketing, model hierarchy or accountants.

No it doesn't. It shows you that the engine can produce more power than with the stock ECU settings, but you're assuming that's the only effect you're getting and that everything else remains equal. What happens with emissions? With fuel efficiency? With safety factors against failure? With the projected lifetime of the engine?

Edit: Here's a paper about fatigue analysis in connecting rods. Note especially how they calculate a predicted number of cycles to failure.

https://iopscience.iop.org/article/10.1088/1757-899X/485/1/012008/pdf

 

[PowerJunkie]

Trying to downshift to 2nd in the AMG @ CGV and you have clicked the paddle 3 times......

 

[pikietana]

This is really stupid, redline has nothing to do with when downshifting is allowed. So many missed shifts that I deleted whole game. Not fun anymore.

 

[eran0004]

This is really stupid, redline has nothing to do with when downshifting is allowed. So many missed shifts that I deleted whole game. Not fun anymore.

It's arguably better this way, since it prevents players from using an unrealistic technique to gain an advantage.

 

[PowerJunkie]

It's arguably better this way, since it prevents players from using an unrealistic technique to gain an advantage.

No, it's not. It's not consistent. There are times the shifts WILL go in and hit the rev limiter, usually accidental shifts for me. They have the parameters wrong, no matter how a lot of you want to argue it. I understand why they are doing it, I agree with that part. The almost instantaneous shifts from 5th to 1st, but they wound this up way too tight and it is not reliable. It NEEDS updated and loosened up a bit.

 

[eran0004]

No, it's not. It's not consistent. There are times the shifts WILL go in and hit the rev limiter, usually accidental shifts for me. They have the parameters wrong, no matter how a lot of you want to argue it. I understand why they are doing it, I agree with that part. The almost instantaneous shifts from 5th to 1st, but they wound this up way too tight and it is not reliable. It NEEDS updated and loosened up a bit.

Perhaps it's not targeted at the rev limit, but slightly above it. So the fact that you sometimes hit the rev limit and sometimes your shift is not accepted doesn't have to mean that it's inconsistent, it might just mean that you're shifting close to the limit and sometimes you're on the wrong side of it.

 

[PowerJunkie]

Perhaps it's not targeted at the rev limit, but slightly above it. So the fact that you sometimes hit the rev limit and sometimes your shift is not accepted doesn't have to mean that it's inconsistent, it might just mean that you're shifting close to the limit and sometimes you're on the wrong side of it.

Sorry, but not how that works, if it is set at an rpm, then it should got into gear at that rpm or lower, but that is not the case. It's not consistent per car, and each car is different. Have you tested this or just speaking on it?

 

[Voodoovaj]

It's arguably better this way, since it prevents players from using an unrealistic technique to gain an advantage.

They used a hammer to turn a screw. Their fix makes the game feel broken. I call for a shift and I get nothing.

Addressing an unrealistic advantage is better addressed (and more effectively addressed) by breaking the engine, which is a feature already in the game. You shift too aggressively, and BOOM, your lap time is done. In addition, as can be seen with the new Porsche, too much rear breaking (which is what the downshift is causing) SHOULD cause the rear to slip as if there is no ABS, which again, would throw your lap out the window.

You can still aggressively downshift by spamming the button (or lever), which unfortunately works better than trying to time the downshifts, so they have absolutely made it worse...MUCH worse.

 

[eran0004]

Sorry, but not how that works, if it is set at an rpm, then it should got into gear at that rpm or lower, but that is not the case. It's not consistent per car, and each car is different. Have you tested this or just speaking on it?

What do you mean? Do you have an example?

 

[PowerJunkie]

What do you mean? Do you have an example?

I will take that as a no, you haven't. I have played/tested this, do I have video/replay proof with a full analysis. No. So my question is, if you haven't, do you use a MT or an AT, going to guess AT. Not that it really matters, as I said before, the group supporting this are the ones it doesn't affect negatively. I will however still stand in what I have said before. I support the idea PD is trying to do, it needed done, but they broke it, and should wind the parameters that they decided to use back just a bit.

 

[NevilleNobody]

I understand why they are doing it, I agree with that part. The almost instantaneous shifts from 5th to 1st, but they wound this up way too tight and it is not reliable. It NEEDS updated and loosened up a bit.

This is my take. I have complained about "gear spamming" in the past so all for the change but too often I go to change down and ..... nothing...... and i do not feel i am being super aggressive at all.

 

[benjoi84]

This isn't an issue that has ever concerned me.

However, if it's true that the motivation for its implementation was to prevent players getting an "unrealistic" advantage over others in the online community, then what the hell are PD doing about players who gain an advantage by full on ruining other's races and crashing into them?!

 

[eran0004]

I have played/tested this, do I have video/replay proof with a full analysis. No.

I didn't ask for video proof. I asked you what you mean when you say it's inconsistent. No need to get overly defensive about it.

 

[PowerJunkie]

I didn't ask for video proof. I asked you what you mean when you say it's inconsistent. No need to get overly defensive about it.

Okay.... so the current CGV time may be able to provide an example. I will try to purposely cause this later tonight. Sector one, the car will drop to second and tap the chip, but in the start of the second sector, going into the left turn, it absolutely refuses to go into second when needed. I just changed what I am doing now, and coasting into the turn in 3rd. I have lost too many 28.4xx sectors because of this. Anyways, I will get a video of it if it happens. If it does it we should see what I mean.
Another example that burns in my head is the Toyota Supra at Grand Valley a couple weeks ago, it would do the same, would not go into second when I 'needed' it to, so I had to adjust, BUT, entering into the tunnel right when the last sector starts, trying to get into second a few times, multiple clicks of the paddle, it would hit 2nd, and almost like it was skipping 2nd and bang into 1st and be over revving like the update had not happened. (killed the lap) That happened more than I like to think about.
That is what I mean, it cannot be RPM based alone, or it would not hit the chip in every scenario. There are time it does, and others it doesn't, in the same car.

 

[PowerJunkie]

Okay...just got done. Made a video with three clips/examples.....what I wrote in the description...
A brief video showing how inconsistent the down shifting is now that Polyphony Digital did and update to the downshifting to eliminate an exploit. I do understand what they are trying to accomplish, but to steal a phrase another player said, they 'fixed' this with a hammer instead of a screwdriver. The first clip shows trying to downshift literally as fast as possible, multiple clicks of the paddle. Very slow timed shifts, rev limiter not engaged a single shift. The second and third clip shows how the downshift will take in an unwanted circumstance and go up high enough to hit the rev limiter and break the rear tires loose. This is the problem, there is no predictability in the down shifting.....it's a crap shoot, you might get the down shift, you might not.

 

[jrbabbitt]

This is why you cannot downshift at high revs, if GT7 could program actual damage a lot of players would have to buy replacement motors....

 

[eran0004]

A brief video showing how inconsistent the down shifting is now that Polyphony Digital did and update to the downshifting to eliminate an exploit. I do understand what they are trying to accomplish, but to steal a phrase another player said, they 'fixed' this with a hammer instead of a screwdriver.

Thanks for sharing the video

The first clip shows trying to downshift literally as fast as possible, multiple clicks of the paddle. Very slow timed shifts, rev limiter not engaged a single shift. The second and third clip shows how the downshift will take in an unwanted circumstance and go up high enough to hit the rev limiter and break the rear tires loose. This is the problem, there is no predictability in the down shifting.....it's a crap shoot, you might get the down shift, you might not.

What I see in these clips is what you describe as "trying to shift as fast as possible". You are redlining the engine with every shift except for the shift into second gear at the last clip. But the goal here isn't to shift down as fast as possible, it's to select a good gear for the corner exit. Sure, shifting down early gives you a little more engine braking but as you observed that could also lead to loss of traction, especially at the lower gears and while cornering.

If you want to shift down early or in rapid succession you need a gearbox with close ratios. That way the speed difference between the gears is not that great. With tall ratios you need to wait for the revs to drop further before you can shift gears, otherwise you will overspeed the engine.

The shifts look very consistent to me (looks like a difference of a few hundred rpm at most, which may be attributable to the timing of the shift), except for the late shift into second gear at the last clip. I'll see if I can replicate that.

 

[jrbabbitt]

Going back to the origin of this condition... PD had discovered that an unfair advantage was being used as players were downshifting at high RPMs as engine braking to slow down quicker than brakes alone which is not fair.

So PD altered the programming to prevent downshifting at high RPM.