I want to see a proper modern sports CVT developed!
Discussion
I know CVTs struggle with high torque outputs, but if a manufacturer were to invest into the R&D I think it would be mint to have a high revving car with a CVT that simulates a DCT.
You might ask: why? Well, first of all, whilst a CVT can constantly vary the gear ratio to sit at the engine's specific peak power RPM - there's a major overlooked component of doing so - and that's drive inneficiency from the belt slipping. Geared right, it's actually better and easier to have a CVT operate within a group of narrow clearly defined stepped gear ratios for performance purposes rather than continuously varying the gear ratio to maintain a specific RPM, since belt slip will reduce the engine power output by ~10-15%.
On the other hand, a CVT is smaller and lighter than a DCT or typical AT - and you don't need to add any parts, weight, or size to give you more or less gears/speeds - or to change the ratios of each of those individual gears/speeds. That mean that you can have a frickin 6, 8, 10, or 12 speed by changing some numbers in some basic code. So, lighter, easier to package, and more efficient (shorter ratios = better for MPG and acceleration).
Not to mention that not only does this get rid of the hideous noise/driving characteristics typically associated with a CVT, but it actually enhances the type of driving characteristics most people have come to love about autos with their lightning fast gear changes and shorter gear ratios.
So, in the end, what would this mean? Well, you'd be able to hear the car scream so much more often, and with more intensity too (thanks to shorter gear ratios that allow the engine to climb the rev counter far easier) and you'd have a better performing car.
I mean, come on, imagine a GT3 or 458 with a 10 speed. And you'd save a bit of weight by giving it that 10 speed too. I saw a video of the auto Subaru WRX and the shift times seemed okay too (not sure how it compares to something like the PDK in actual milliseconds though) - but I guess it doesn't even matter since power delivery continues uninterupted with a CVT regardless.
What you guys think? Lol.
You might ask: why? Well, first of all, whilst a CVT can constantly vary the gear ratio to sit at the engine's specific peak power RPM - there's a major overlooked component of doing so - and that's drive inneficiency from the belt slipping. Geared right, it's actually better and easier to have a CVT operate within a group of narrow clearly defined stepped gear ratios for performance purposes rather than continuously varying the gear ratio to maintain a specific RPM, since belt slip will reduce the engine power output by ~10-15%.
On the other hand, a CVT is smaller and lighter than a DCT or typical AT - and you don't need to add any parts, weight, or size to give you more or less gears/speeds - or to change the ratios of each of those individual gears/speeds. That mean that you can have a frickin 6, 8, 10, or 12 speed by changing some numbers in some basic code. So, lighter, easier to package, and more efficient (shorter ratios = better for MPG and acceleration).
Not to mention that not only does this get rid of the hideous noise/driving characteristics typically associated with a CVT, but it actually enhances the type of driving characteristics most people have come to love about autos with their lightning fast gear changes and shorter gear ratios.
So, in the end, what would this mean? Well, you'd be able to hear the car scream so much more often, and with more intensity too (thanks to shorter gear ratios that allow the engine to climb the rev counter far easier) and you'd have a better performing car.
I mean, come on, imagine a GT3 or 458 with a 10 speed. And you'd save a bit of weight by giving it that 10 speed too. I saw a video of the auto Subaru WRX and the shift times seemed okay too (not sure how it compares to something like the PDK in actual milliseconds though) - but I guess it doesn't even matter since power delivery continues uninterupted with a CVT regardless.
What you guys think? Lol.
The modern 8/9/10 speed torque converter gearboxes pretty much act as CVTs anyway, with very little between the gears at road speeds, and bouncing between tickover and redline if you are binary on the throttle.
Perhaps the gap in the market is to have a CVT with fewer fake gears (eg. 4 or 5) so you can have the enjoyment of accelerating through the rev range in a car with a tuneful engine.
Perhaps the gap in the market is to have a CVT with fewer fake gears (eg. 4 or 5) so you can have the enjoyment of accelerating through the rev range in a car with a tuneful engine.
Torotrak have been doing advanced CVTs for some time, amongst others no doubt. Look up their toroidal variator CVT, which uses an electronically controlled ratio arrangement.
The original purpose for this technology if I recall was to use it to drive a flywheel as a means of storing kinetic energy in vehicles.
Some of the KERS systems, for both racing and road cars, that have been examined over the last couple of decades have been based around flywheel storage, which normally has a rotor of relatively large inertia spinning in a vacuum at fairly ridiculous speeds.
The CVT is used to change the flow of energy in and out of the flywheel, as it can vary its ratio by a factor of 10 I believe.
The Torotrak variator was also proposed as a means of down-sizing electric traction motors for EVs.
The high torque density of state-of-the-art traction motors, combined with higher and higher operating speeds for these has meant that a simple fixed final drive ratio is usually the chosen route for EV applications. In some cases, 2 -speed gearboxes are used.
As for belt-Iess CVTs I was involved in designing something along these lines many years ago, using an epicyclic gearbox. The idea was to use an electric motor to control the speed of what would normally be the stationary element, normally the ring gear or planet carrier. By forcing the stationary element to rotate in either direction it varies the ratio between the other two rotating elements, and being a geared system, there is no slip possible, so the efficiency is determined by the normal gear losses and the power to drive the electric motor. My guess is that it was also considered by the likes of Torotrak and they probably found that the toroidal variator was a superior solution as it has neither belts nor gears, and doesn't require power to drive an electric motor.
Anyway, such gearboxes already exist, my guess is that they don't give the driving feel that you are looking for, even when configured as suggested with stepped ratios. Having said that, I'd also suggest that something like the variator is probably torque-limited for a given physical size since it relies on rolling friction rather than gear teeth. In which case, maybe the motor-controlled conventional epicyclic I described above would give you the torque rating and ratio range you are looking for.
Truth is, a conventional auto with torque converter is probably going to be the right answer...
The original purpose for this technology if I recall was to use it to drive a flywheel as a means of storing kinetic energy in vehicles.
Some of the KERS systems, for both racing and road cars, that have been examined over the last couple of decades have been based around flywheel storage, which normally has a rotor of relatively large inertia spinning in a vacuum at fairly ridiculous speeds.
The CVT is used to change the flow of energy in and out of the flywheel, as it can vary its ratio by a factor of 10 I believe.
The Torotrak variator was also proposed as a means of down-sizing electric traction motors for EVs.
The high torque density of state-of-the-art traction motors, combined with higher and higher operating speeds for these has meant that a simple fixed final drive ratio is usually the chosen route for EV applications. In some cases, 2 -speed gearboxes are used.
As for belt-Iess CVTs I was involved in designing something along these lines many years ago, using an epicyclic gearbox. The idea was to use an electric motor to control the speed of what would normally be the stationary element, normally the ring gear or planet carrier. By forcing the stationary element to rotate in either direction it varies the ratio between the other two rotating elements, and being a geared system, there is no slip possible, so the efficiency is determined by the normal gear losses and the power to drive the electric motor. My guess is that it was also considered by the likes of Torotrak and they probably found that the toroidal variator was a superior solution as it has neither belts nor gears, and doesn't require power to drive an electric motor.
Anyway, such gearboxes already exist, my guess is that they don't give the driving feel that you are looking for, even when configured as suggested with stepped ratios. Having said that, I'd also suggest that something like the variator is probably torque-limited for a given physical size since it relies on rolling friction rather than gear teeth. In which case, maybe the motor-controlled conventional epicyclic I described above would give you the torque rating and ratio range you are looking for.
Truth is, a conventional auto with torque converter is probably going to be the right answer...
Personally I think Honda have nailed it with the new hybrid Civic. A Type R version would be brilliant and have a feeling this is going to be the hybrid drive for the next model.
Rory Reid managed to get 80 mpg on a road test, which has never really been achieved in a Prius with a conventional CVT. That's how I'd like a future hot hatch. Economic in regular driving but plenty of power when you need it. Sadly no gearbox but I can live with that.
There's a mass penalty but think you could engineer this out a bit. The regular Civic is about 1500kg with no mass saving, but given the new M2 is 1700+, that's not bad at all by modern standards.
Rory Reid managed to get 80 mpg on a road test, which has never really been achieved in a Prius with a conventional CVT. That's how I'd like a future hot hatch. Economic in regular driving but plenty of power when you need it. Sadly no gearbox but I can live with that.
There's a mass penalty but think you could engineer this out a bit. The regular Civic is about 1500kg with no mass saving, but given the new M2 is 1700+, that's not bad at all by modern standards.
georgeyboy12345 said:
Are manufacturers going to bother with any of this any more given electric vehicles are the future?
Many mainstream manufacturers have already said they have ceased ICE development already. You'd assume that covers gearboxes etc as well. What the Getrags and ZFs of the world do at this point I don't know... I guess EVs usually have a reduction gearbox, so maybe they switch to that?
Truckosaurus said:
The modern 8/9/10 speed torque converter gearboxes pretty much act as CVTs anyway, with very little between the gears at road speeds, and bouncing between tickover and redline if you are binary on the throttle.
Perhaps the gap in the market is to have a CVT with fewer fake gears (eg. 4 or 5) so you can have the enjoyment of accelerating through the rev range in a car with a tuneful engine.
Yeah I get what you're saying, but they're also heavier and much larger. Not to mention that they can still suffer from being too 'slushy' compared to an actual DCT depending on how it's setup/tuned.Perhaps the gap in the market is to have a CVT with fewer fake gears (eg. 4 or 5) so you can have the enjoyment of accelerating through the rev range in a car with a tuneful engine.
CrgT16 said:
I always thought on paper CVTs are good as take the best rpm range of the engine. In practice they are not the best by reasons you mention.
Gearboxes are reaching peak development with the downfall of ICE development.
Cars are going from VHS to CD.
Unfortunately, you're right about that. However, I do remain optimistic. EVs still have issues to iron out, and with cleaner fuels the ICE regains a lot of merit. Gearboxes are reaching peak development with the downfall of ICE development.
Cars are going from VHS to CD.
Kawasicki said:
Doesn’t belt slip occur even when in a set gear?
So a stepped Lay-out also suffers from efficiency loss.
I think that's if the belt can't handle the torque - but I haven't heard of that being a standard operating principle of the CVT in general. Once it's in a stepped gear there shouldn't be any slip.So a stepped Lay-out also suffers from efficiency loss.
Zero Fuchs said:
Personally I think Honda have nailed it with the new hybrid Civic. A Type R version would be brilliant and have a feeling this is going to be the hybrid drive for the next model.
Rory Reid managed to get 80 mpg on a road test, which has never really been achieved in a Prius with a conventional CVT. That's how I'd like a future hot hatch. Economic in regular driving but plenty of power when you need it. Sadly no gearbox but I can live with that.
There's a mass penalty but think you could engineer this out a bit. The regular Civic is about 1500kg with no mass saving, but given the new M2 is 1700+, that's not bad at all by modern standards.
Well, with a conventional CVT (like Subaru's lineartronic transmission), I think the mass penalrty isn't really there. Still heavier than the manual transmission they offer with the WRX, but lighter and smaller than comparative DCTs or torque converter ATs I believe. Rory Reid managed to get 80 mpg on a road test, which has never really been achieved in a Prius with a conventional CVT. That's how I'd like a future hot hatch. Economic in regular driving but plenty of power when you need it. Sadly no gearbox but I can live with that.
There's a mass penalty but think you could engineer this out a bit. The regular Civic is about 1500kg with no mass saving, but given the new M2 is 1700+, that's not bad at all by modern standards.
But, yeah, cars are getting way too big and heavy. It's absolutely ridiculous. Do everything they can to improve fuel emissions and then add 200kg to the weight of the car lol.
GT9 said:
Torotrak have been doing advanced CVTs for some time, amongst others no doubt. Look up their toroidal variator CVT, which uses an electronically controlled ratio arrangement.
The original purpose for this technology if I recall was to use it to drive a flywheel as a means of storing kinetic energy in vehicles.
Some of the KERS systems, for both racing and road cars, that have been examined over the last couple of decades have been based around flywheel storage, which normally has a rotor of relatively large inertia spinning in a vacuum at fairly ridiculous speeds.
The CVT is used to change the flow of energy in and out of the flywheel, as it can vary its ratio by a factor of 10 I believe.
The Torotrak variator was also proposed as a means of down-sizing electric traction motors for EVs.
The high torque density of state-of-the-art traction motors, combined with higher and higher operating speeds for these has meant that a simple fixed final drive ratio is usually the chosen route for EV applications. In some cases, 2 -speed gearboxes are used.
As for belt-Iess CVTs I was involved in designing something along these lines many years ago, using an epicyclic gearbox. The idea was to use an electric motor to control the speed of what would normally be the stationary element, normally the ring gear or planet carrier. By forcing the stationary element to rotate in either direction it varies the ratio between the other two rotating elements, and being a geared system, there is no slip possible, so the efficiency is determined by the normal gear losses and the power to drive the electric motor. My guess is that it was also considered by the likes of Torotrak and they probably found that the toroidal variator was a superior solution as it has neither belts nor gears, and doesn't require power to drive an electric motor.
Anyway, such gearboxes already exist, my guess is that they don't give the driving feel that you are looking for, even when configured as suggested with stepped ratios. Having said that, I'd also suggest that something like the variator is probably torque-limited for a given physical size since it relies on rolling friction rather than gear teeth. In which case, maybe the motor-controlled conventional epicyclic I described above would give you the torque rating and ratio range you are looking for.
Truth is, a conventional auto with torque converter is probably going to be the right answer...
Wow, thanks for the insight! You seem quite knowledgable haha. Yeah, i feel like those epicyclic-style CVTs defeat the main purpose of being lighter, smaller, and less complicated - but for sure it's probably easier for manufacturers to 'make them work' considering both Honda and Toytoa have leaned towards them. The original purpose for this technology if I recall was to use it to drive a flywheel as a means of storing kinetic energy in vehicles.
Some of the KERS systems, for both racing and road cars, that have been examined over the last couple of decades have been based around flywheel storage, which normally has a rotor of relatively large inertia spinning in a vacuum at fairly ridiculous speeds.
The CVT is used to change the flow of energy in and out of the flywheel, as it can vary its ratio by a factor of 10 I believe.
The Torotrak variator was also proposed as a means of down-sizing electric traction motors for EVs.
The high torque density of state-of-the-art traction motors, combined with higher and higher operating speeds for these has meant that a simple fixed final drive ratio is usually the chosen route for EV applications. In some cases, 2 -speed gearboxes are used.
As for belt-Iess CVTs I was involved in designing something along these lines many years ago, using an epicyclic gearbox. The idea was to use an electric motor to control the speed of what would normally be the stationary element, normally the ring gear or planet carrier. By forcing the stationary element to rotate in either direction it varies the ratio between the other two rotating elements, and being a geared system, there is no slip possible, so the efficiency is determined by the normal gear losses and the power to drive the electric motor. My guess is that it was also considered by the likes of Torotrak and they probably found that the toroidal variator was a superior solution as it has neither belts nor gears, and doesn't require power to drive an electric motor.
Anyway, such gearboxes already exist, my guess is that they don't give the driving feel that you are looking for, even when configured as suggested with stepped ratios. Having said that, I'd also suggest that something like the variator is probably torque-limited for a given physical size since it relies on rolling friction rather than gear teeth. In which case, maybe the motor-controlled conventional epicyclic I described above would give you the torque rating and ratio range you are looking for.
Truth is, a conventional auto with torque converter is probably going to be the right answer...
It's just a shame imo because the only real limiting factor is being able to relibaly handle more torque. If that problem could be solved then i'm sure it would be as simple as programming them a certain way.
Baldchap said:
Zero Fuchs said:
Rory Reid managed to get 80 mpg on a road test, which has never really been achieved in a Prius with a conventional CVT.
You might want to tell the Prius owners they can't get 80mpg... 
It's know instead as a power split device or PSD (Hybrid Synergy Drive if you want the marketing term) and is based on a planetary gearset design as mentioned above by GT9.
Toyota/Lexus do have some way of mimicking distinct gears with some versions of this system. Don't know how they do it mind.
Fairly sure the Honda hybrid system isn't quite as good as the Toyota one. Based on RAV4 hybrid Vs CRV the RAV4 was still able to clear several MPG more than the CRV. Could be down to the type of car though. It may be better on smaller vehicles.
Edited by Otispunkmeyer on Wednesday 5th April 18:15
Edited by Otispunkmeyer on Wednesday 5th April 18:16
GT9 said:
Torotrak have been doing advanced CVTs for some time, amongst others no doubt. Look up their toroidal variator CVT, which uses an electronically controlled ratio arrangement.
I was working at Ricardo's when this was being proposed. It's basic problem it weighed about 2-3 times more than the engine it was coupled to. It eventually went bust.
robinessex said:
GT9 said:
Torotrak have been doing advanced CVTs for some time, amongst others no doubt. Look up their toroidal variator CVT, which uses an electronically controlled ratio arrangement.
I was working at Ricardo's when this was being proposed. It's basic problem it weighed about 2-3 times more than the engine it was coupled to. It eventually went bust.Funnily enough I remember an exchange of posts between you and I many years ago about a different type of multi-ratio gearbox technology, I can't recall exactly what it was.
What I do remember was you asking about using its applicability to electric motors to increase torque at the driven wheels from a lighter motor.
The response I gave was that its cheaper and easier to just add a modest amount of extra copper and iron into the motor, and/or increasing the speed rating.
The higher speed rating allows fo a higher (fixed) final-drive ratio, thus increase torque at the driven wheels from a given motor size.
Not to mention more efficient and more reliable.
Basically, it doesn't usually pay to introduce a multi-ratio system, the mass penalty is far higher than just putting the effort into the motor itself.
Some high-performance EVs have a two-speed gearbox, but the vast majority have the fixed ratio final drive set-up.
The wide speed capability is managed in the motor drive, aka 'electronic gearbox'.
Otispunkmeyer said:
Prius doesn't have a CVT anyway. Not in a conventional sense.
It's know instead as a power split device or PSD (Hybrid Synergy Drive if you want the marketing term) and is based on a planetary gearset design as mentioned above by GT9.
Toyota/Lexus do have some way of mimicking distinct gears with some versions of this system. Don't know how they do it mind.
Fairly sure the Honda hybrid system isn't quite as good as the Toyota one. Based on RAV4 hybrid Vs CRV the RAV4 was still able to clear several MPG more than the CRV. Could be down to the type of car though. It may be better on smaller vehicles.
The CRV uses an older version of Honda’s EHEV. The latest Civic is miles better and has moved ahead of Toyota in my opinion. It's know instead as a power split device or PSD (Hybrid Synergy Drive if you want the marketing term) and is based on a planetary gearset design as mentioned above by GT9.
Toyota/Lexus do have some way of mimicking distinct gears with some versions of this system. Don't know how they do it mind.
Fairly sure the Honda hybrid system isn't quite as good as the Toyota one. Based on RAV4 hybrid Vs CRV the RAV4 was still able to clear several MPG more than the CRV. Could be down to the type of car though. It may be better on smaller vehicles.
Edited by Otispunkmeyer on Wednesday 5th April 18:15
Edited by Otispunkmeyer on Wednesday 5th April 18:16
Gassing Station | General Gassing | Top of Page | What's New | My Stuff