50 replies to this topic

[Ray Bell]

Because I don't want to go wading through months of topics, and because it may have been wiped anyway, I'm starting a new thread with this picture:



This is the constantly variable transmission unit developed by DAF (now part of Volvo) and used in a Brabham F3 car circa 1966.

[Marco94]

Ray,

May I be the first to congratulate you on your excellent choice of subject. I am an engineer by training and the CVT is the only true transmission. Gearboxes are just a very primitive device, used because people couldn't build strong enough CVT's after the first ten years of car development. I have seen one of the F3 cars drive through the inner city of Eindhoven, where DAF is still located. My parents have been driving cars with CVT for the past 25 years or so, so I absolutelly love the CVT. They are being used in more and more cars these days. Top examples are the Rover MGF and Audi A6!

The story on DAF is a little more complicated. Called Van Doornes Automobiel Fabriek, it designed and build both passenger cars (the wellknown "trutte schudders") and trucks. The passenger cars division, has indeed been absorbed in or named Volvo Car. I am not sure which is correct. The truck division had been very succesfull until 1993 when it went broke. After theat, the truck division has been build up again and is now part of Pacar.

You can find some more pictures and text at: http://www.ritzsite....AF_cars_p17.htm and at http://www.dafclub.nl/.

Marco.

[desmo]

Ah, back to CVTs! Proscribed by the FIA's maximum of 7 forward speeds in F1. [Insert stock rant on F1's idiotic technical regs]

[Marco94]

Funny thing is that it is possible to use a CVT in Formula 1! The MGF has the answer. Actually Porsche had it first. OK the explenation:

Porsche started building cars with automatic gearboxes around the time they were racing their PDK double clutch system. They did so by introducing the Tiptronic. This devise has two settings. One is automotic, but the other is a sequential like manual up and down shift setting. The driver can therefor choose between auto or manula shifting. That is the big marketing trick. You see, 99% of the drivers use automatic shifting and don't use the manual option at all.

MG used the same trick. You can use their CVT continuesly variable or with the manual option, with F1-like steer shift buttons. When using the manual option you have a virtual 5-speed transmission.

This manual option could be used in F1. You can choose anywhere between 1 and 7 speed ratios, using the same transmission. You don't need to disengage the clutch and shifts are much, much quicker then even the current F1 gearboxes. I don't know if it brings a big advantage, but it is an interesting technical project and the image of CVT's wouldn't be hurt either.

Marco.

[desmo]

Yes, but what a waste to use a CVT that way! A true "stepless" CVT that could handle the torque loads of an F1 motor, and the Audi A6 version is as close as I've seen yet, would open the door for a much less comprimised engine as far as tuning goes. Transient response out of corners could be much sharper as one wouldn't have to corner in the dead zone below the power curve any longer. I can see a TC (while we're ignoring the rulebook) where torque would be modulated not by throttle or timing but by the gearbox using input from the ECU. I'll bet that with a given horsepower, one could lap quite a bit quicker this way. And the benefits would spin off into road cars as increased safety, performance and fuel economy (although I imagine the sounds would be strange)! But alas...[stock rant against idiotic FIA tech regs]

[Ray Bell]

des, I tried to post that 'stock rant,' but the URL was rejected!

[Marco94]

Des,

You are absolutly right in feeling this way about the FIA rules. I was just making the point that if the engineers wanted, the _could_ get real race experience with the transmission system of a CVT eventhough it is used stepped.

Marco.

[Ray Bell]

For efficiency, it's obviously a great thing, but for driving, well, I like changing gear, H-pattern and all. I prefer six speeds, but grabbing top gives me a grating sound all too often... must be because I've only got a five-speed!

[Yelnats]

Ban the bloody things! {;-> Only half joking of course! But honestly, I have watched several races featuring constant velocity transmisions in Canada and you have never heard such a mournful wailing as these machines make. The vehicles I am referring to are a mini-NASCAR type using 135 hp engines derived from snowmobile engines.

The sound I am reffering to is the wa-wa-wail these things make as the tranny responds to every touch of the throttle by varying the engine speed erratically up and down the register. When 20 of these vehicles pass it's like listening to a pack of love-sick cats!

I can't see any major racing series surviving long with such a pussy-cat exhaust note ! {;->

[Ray Bell]

I would have thought to work properly they would have kept the engine revs pretty constant, but then, it's a throttle position thing, isn't it?

[PDA]

Neither the Porsche nor the MG have CVT transmission, both use a number of fixed gear ratios, plus a hydraulic torque converter

[Ray Bell]

I thought the tiptronic arrangement was more like that... thanks. I doesn't have the standard kind of epicyclic gears, though, does it?

[GoAlesi]

Yelnats, :lol: I imagine it would sound very much like so many wailing cats!

Heres an extract from an article by LJK Setright no less:
" 2 or 3 years ago, the latest evolution in the story of the Van Doorne's transmission was built into a Williams F1 racing car....Driven by young D Coulthard, the Williams lapped Silverstone two clear seconds faster than a car with conventional gearbox. The CVT that would have made it possible for even a frail old lady to drive this car also enabled young Coulthard to drive faster than ever before." CAR March 1997

The article was entitled "Going Nowhere Fast" and basically is a rant about how restrictive the F1 technical regulations are. It is hilarious in places, for instance Setright points out that bodywork braced by tension wires was old-fashioned in aviation way back in the 1920s.

How efficient are CVTs? I mean isn't there some energy loss due to slippage in the belt drive, and does this place a limit on the amount of torque that can be used with a CVT? I guess if Williams got a CVT to work with a 700+hp F1 engine it can't be that much of a problem.

[desmo]

I would love to read that article by Setright. In fact, I highly recommend reading anything you can get your hands on by LJK. He would be the ultimate poster for this forum (except maybe a certain Mr. Newey!).

[Ray Bell]

Well, I used to live next door to some people named van Doorne... came here in the great migratory push of the fifties.
No relation? Well, I bet they never drove a DAF, either, but were bound to have had a Holden some time.

[Marco94]

PDA,

Just to clarify things: Porsche's Tiptronic does not use a CVT. You are right about that, but then I never said they did use a CVT! As for the MG F Steptronic, see for yourself: http://www.mgcars.or...mg/changing.htm.

Marco.

[Ray Bell]

Shame there are no cutaways or explanations of the workings.

[desmo]

Heres a good desription I found of Honda's CVT:




The continously variable transmission operates in a fashion unlike any other conventional transmission.
Instead of gears, a steel belt is used to transfer power across variable-sized pulleys. Where a conventional
transmission makes use of a limited number of gear ratios, the CVT can access an infinite number or drive
ratios. It sounds complicated, but the mechanicals are actually simpler than those used in conventional
transmissions. This page is inteneded to explain some of the mechanicals.

The most interesting part of this transmission is the varible-diameter pulleys that are used.
"Variable-diameter" is actually a misleading term to use, since it is the active position of the belt that
changes on the pulley, not the actual pulley diameter. Each pully consists of two cone-shaped halfs mounted
on the same axis facing each other. When viewed from the side, there is a V-shaped gap in the middle. This
is where the belt rides.

The belt is made from steel mesh, and is V-shaped to fit the gap of the pulley. The two halfs of the pulley can
be moved closer together or farther apart, changing the width of the gap. When the halfs are farthest apart,
the belts rides at the "bottom" of the V, which is the minimum radius. As the halfs are pushed closer together,
the belt is forced farther out of the gap, and rides at a greater radius. In this manner, the effective radius of
the pulley is changed by moving the pulley halfs relative to each other.

One of these pulleys is connected to the engine, the other is connected to the drivetrain heading out to the
front wheels. One pulley is moved to control the drive ratio, while the other pulley maintains the proper gap to
tension the belt. The ratio of these pulley diameters corresponds to the gear ratios of different gears in a
conventional transmission.

When the car comes to a complete stop, the drivetrain must be disconnected. This is done by an automatic
multi-plate clutch. While the clutch is disengaged, the drivetrain is still partially linked to the engine to provide
a slight "forward creep" when you let off the brake at a standstill. As gas is applied, the engine revs up, the
car moves forward due to the partial coupling, and then the clutch engages just as the engine speed and
drivetrain speed are matched. From that point on, the clutch is firmly engaged until the car comes to a
complete stop. There is no power loss from a torque converter as in a conventional automatic transmission.

The drive-ratio control is the most impressive part of the CVT. Everthing is controlled by the Transmission
Control Module (TCM). The TCM monitors seveal sensors within the car, but most importantly it watches
throttle position, ground speed, and engine rpm. It uses the position of the throttle and speed of the car find
the appropriate engine speed. Engine speeds are stored in a map of values withing the TCM. The target
engine rpm is then compared to the actual observed rpm, and the drive ratio is adjusted to make the figures
match.

Overall, the CVT is simpler and has fewer moving parts than an conventional transmission. In addition, the
existing parts are subject to less abuse than conventional parts. There is never a situation where gear teeth
suddenly come into contact, or the car is jerked by sudden engagement of the drivetrain. The "clutch"
receives very little wear because the it is amlost always engaged. And because of the smooth change of
ratio, the engine and drivetrain are never subjected to any harsh sudden movements. So in theory, the CVT
will be much more durable than a conventional transmission.

Supposedly, there's a limit to the amount of power that can safely be transferred through the CVT. The
current design is intended to perform reliably while handling 130hp or less. The CVT is only mated to the
115hp engine in the United States. And Honda warns that the CVT will experience approximately 30% to
45% more wear for every 2hp added above the 115hp figure.

The CVT is able to keep the engine revving withing a specific range to produce the desired results. The
primary goal of the CVT was to yield high mpg. Therefore, the CVT revs the engine close to its most efficient
operating rpm while the gear selector is in Drive. If you bump it into Sport and mash the gas, the CVT revs
the engine all the way up to peak horsepower for maximum power output. And it stays in the power band the
whole time for impressive acceleration! If you push it further, into Low, the CVT maintains a high drive ratio
and high engine speed that produces a lot of torque at the wheels and extra engine braking.

[PAD]

Yelnats - I think I would have to agree with you regarding the noise. I drove a Ford Escort with CVT and it was really quite odd; from a standing start you put your foot down and the revs immediately went up to 4000rpm and then just sat there until you released the accelerator pressure. The noise was constant, you didn't get that engine note rise that you expect. An F1 race would sound like Heathrow Airport's runway.

On the subject of CVTs, I remember my Grandparents having a DAF 33 (circa 1973 vintage) - in retrospect it looked like a Trabant!! They often had to have the "elastic bands" replaced as they broke every few months (before the modern steel belt). It also had two sets of pulleys/CVTs replacing the need for a differential.

[BT52]

Neither the Porsche or the MGF use a continuously variable transmission.
They are both simply auto's which have convenient up and down buttons near the steering wheel.

[Marco94]

BT52,

Read before you write! I already answered the exact same remark earlier in this thread.

Marco.

[mat1]

Originally posted by desmo
I would love to read that article by Setright. In fact, I highly recommend reading anything you can get your hands on by LJK. He would be the ultimate poster for this forum (except maybe a certain Mr. Newey!).

Yes. Where is LJKS now? is he still writing? Or is he retired?

Mat1

[PDA]

marco - many thanks for that link. I also would like to see a cutaway drawing to be fully convinced. The reason I say this is based on the fuel consumption figures, which show the CVT to be significantly worse than the manual. Now the only CVT car I have driven is the Honda Civic, which accelerates better than a standard Civic, yet has better fuel consumption. Double cone/steel belt CVT is supposed to be at least as efficient as a modern manual gearbox, so one would not expect worse fuel consumption. One would also expect equal acceleration times. Forgive me if I seem skeptical in the abscence of drawings.

[Marco94]

PDA,

The modern push belt CVT as build for instance by VDT, does have an efficiency compareble with two gears in contact as used in manual boxes. Also, a CVT usually does not use a clutch, but a hydro dynamic torque converter. This usually has a negative influence on your fuel consumption, just like stepped automatics. For a good transmission, the fuel consumption should not be more then 2-3 % higher. Anything higher would be driver influence.

A CVT does require a different driving style. I have no idea what it is like to drive a CVT for the first time. My parents drive CVT since 1975 or something like that. What I have heard is that people do have a tendency to floor the throttle in a CVT. Absolutely useless.

When you drive a stepped transmission, you basically control the revs of the engine in order to control the speed. With a CVT all you do is control the torque, by way of a constant throttle opening. I know this from experience. On a flat road for instance, I would be doing 80 kmh. If I then go up a hill, my engine revs (~torque) don't change, but my speed drops. The engine torque reamains constant, but going up hill it should increase. So the CVT "shifts" down. Going down hill the opposite occurs.

Marco.

[IndyIan]

The only problem I can see with CVT in F1 is that with a purely mechanical system(like in a snowmobile) is that when you let off the gas the transmission "gears up" or the pulleys move so drive ratio decreases (goes towards 1:1) So when you hit the gas again there is a slight lag as the pulleys move to increase the drive ratio so the engine can gain rpm's to produce power. The cars would drive like they did in the turbo era when the drivers hit the gas mid-way through the corner so the power would be there on the exit.
With an electronically controlled CVT it would be too easy to put in the evil traction control. Setting the pulleys to change the ratios only as fast as the car can accelerate would result in starting TC.

It would make for interesting engines though. Imagine engines that only had to operate over a 1000rpm range. I think we would see over 1000bhp overnight!

[Ray Bell]

Wouldn't it be easy to set the transmission to change its mode with throttle opening. Then you'd have it 'changing down' as the throttle comes off for a corner, maybe only varying the revs by 500 or so, but still giving some more engine braking, while power on and it resumes drive mode.
Feasible or not?

[IndyIan]

Ray,
I guess it might be possible to have a mechanical system to change the operating mode of the CVT using engine vacuum or a throttle position sensor.

I know for snowmobile motocross they are experts at tuning the clutch and pulley movements so the transmission lag is minimized.

[MrAerodynamicist]

There was a good article about CVT in the free Audi magazine I get [well, my parents get] I@ll see if I can find it and ocr it

[Tigalola]

In the Autocourse publication "Technology Of The F1 Car" it states "the ban was introduced because spectators would find the sound of an entire grid of cars running at full throttle for virtually the whole race monotonous!". If that's how the FIA really feel, why the hell don't they make rules that allow better racing! Gotta agree the sound of those cars would suck.

[MrAerodynamicist]

Okay, finally found that article. Here it is. Sorry for any bits that don't make sense, blame it on the OCR software
--------------------------
Copyright: The Audi Magazine. Issue 23. Summer 2000

DESIGN ANALYSIS
Whatever the conditions. whatever the speed, Audi's multitronic transmission gives you the right gear. Geoff Howard talks to the man behind Audi's new automatic.

REINHARD GESENHAUS, the head of transmission engineering at Audi, is a man with a 9 secret passion. At weekends in the winter he tinkers in his garage with a pair of Harley Davidsons, one a 1942 750cc side-valve V twin WLA 45, the other a new 1340cc Softail (a modern recreation of the same classic design but with an ohv engine and fuel injection). On summer weekends he thunders around the roads of Bavaria on one or the other, cruising in the sunshine like so many other Harley fans. "I imported my ex-military Harley from Russia two years ago," he says "It was one of 40,000 shipped over during the war to provide trans, port for the Allied troops. I've been restoring it ever since.'

Born and educated near Cologne, Dipl-Ing Gesenhaus moved south to join Audi in 1973. Apart from a couple of years working on manual transmissions for Ford, he's been with the company ever since and now has more than a decade as Audi's transmission engineering chief behind him. Like so many top Audi engineers, he's a hands, on guy who loves all things mechanical That's why he talks so proudly of his latest baby, the Audi multitronic. He knows how well it works and he knows precisely why.

It takes more than enthusiasm to design the world's best automatic transmission, though. Indeed, it's taken Audi around ten years of advanced research, more than five years of product development and the full-time employment of 40 engineers - equivalent to 300 man-years of engineering work. In that time the team had to find new materials and lubricants, new ways of interlinking hydraulics and electronics and new standards of manufacturing precision. Along the way it filed between 40 and 50 patents.

The result is a system that provides whatever the driver wants at the flick of a lever. If you're looking for refinement, it can be smooth and shift free, seamlessly matching the gearing to the engine load and speed whatever the conditions. If you're looking for more control, it can be transformed instantly into a sporty close-ratio six-speed sequential gear, box operated by switches on the steering wheel or by the central lever.

"We began working on a continuously variable transmission in the early 1980s," says Herr Gesenhaus. "But for several years it was just an experimental project." CVT has been the powertrain engineer's dream since the dawn of the last century Its ability to constantly match its gear ratio to the driving conditions offers a smooth two-pedal drive with high operating efficiency and good performance. But it relies on the friction of a belt or chain to transmit power from the engine to the wheels, and that, historically, has always been its downfall. Before Audi started on the project, applications were confined to small engines and relatively lightweight cars.

"In the early days there was quite a low limit on the torque we could transmit, so we devised a composite transmission that used manual gear for first and second, plus a variable ratio for third. We called it the '2 plus' system," says Reinhard Cesenhaus. 'Trom the beginning we used a special chain running between two vee shaped pulleys and driving in tension rather than a steel belt pushing metal blocks in compression like other people did. We found the chain could work at a tighter radius and in our installation we needed the most compact design."

By the end of the 1980s, a more advanced 'I plus V' version had bee developed that shifted automatically between a fixed-ratio first gear and variable ratio second. "But it was still unacceptable to most drivers, so w started a completely new project that was extremely close to the concept we have today. We designed it primarily to improve fuel economy in response to rising fuel prices. After testing ten prototype transmissions on rigs and in cars to prove the concept worked, we began pre-program development of a production design.

The Audi multitronic doesn't need the epicyclic gear and power absorbing torque converter of a conventional automatic transmission. A multi-plate wet clutch engage automatically when moving off and two vee-shaped pulleys transmit drive using a special link-plate steel chain. This acts just like a rubber vee-belt - transmitting torque between two pulleys by friction. But it is many times stronger than a rubber belt, free from stretch an wear, and virtually silent in operation.

The pulleys too are very special. Each is designed to fit together as separate halves, sliding on a common hub so the width of the vee groove can be opened or close under hydraulic pressure. This causes the chain to rid up or down the vee between the two halves. At the bottom of the groove, it runs on a tight radius and at the top end on a large radius. Since the chain length is fixed, on pulley closes as the other opens, varying the gearing ratio continuously over an exceptionally wide range - a spread of over 6.1 in the Audi design compared with just over 5.0 for a conventional five-speed automatic.

One of the most vital elements in the operation of this advanced design is the hydraulic control system that adjusts and holds the pulley halve together. It must provide sufficient clamping loads to prevent slip and the same time respond immediately to change the ratio whenever the need arises, either by increasing pressure to close the groove width of the driven pulley for lower gearing, or increasing pressure to close the groove on the driving pulley to provide higher gearing. The balance in the pressure is controlled internally by a mechanical/hydraulic system. "We call it a self-regulating torque sensor,' says Herr Gesenhaus. "It works like a pressure limiting valve to detect the torque flowing to the transmission from the engine and then establish equilibrium with the contact force, preventing slip and eliminating any losses that would be caused by over-clamping. 1 responds extremely quickly to sudden inputs that could cause momentary slip. Otherwise we would need to raise the basic clamping pressure to cove these conditions, which would increase the hydraulic losses."

Only the ends of the pins between the chain links make contact wit the grooves. The development of their design was crucial to the system's success. "First, they had to be strong and durable, to transmit maximum torque throughout the vehicle life without fatigue," he says. "Then we had to develop exactly the right finish and lubricant for the contact surfaces. They can't he too smooth, or there won't be enough friction to bite with, out slip. And they can't be too rough, or there will be too much wear in use. We use ball bearing steel for the pins and gearwheel steel for the pulleys, both with specific heat treatment that provides a casehardened surface.

Working with Burmah Oil we developed a unique synthetic transmission fluid that also lubricates the wet clutch operates the hydraulics. Even though there are only nine pins transmitting all the drive when the chain is at its tightest radius, they can operate for the whole life of the transmission without wearing more than one or two tenths of a millimetre.'

The hydraulic system is operated by a sophisticated electronic controller that is integrated with the transmission and connected to the engine management module by a highs-peed data bus. It monitors vehicle operation and stores data on the driving le so it can dynamically adapt to the conditions and the way the cars being driven, for the best possible response. With the extra flexibility of the multitronic transmission, it is even more intelligent than the DSP (Dynamic Shift Program) provided on other Audi automatics. It adjusts the way the car moves off from rest according to how hard the diver opens the throttle, controls the degree of creep during parking manoeuvres regardless of engine temperature and even reduces drag in traffic whenever the brakes are being applied.

One of the main advantages of this driver recognition program (DRP) is its ability to track the vehicle speed and eliminate the so-called 'rubber band effect of other CVTS. "We did a lot of work in development to provide an acceptable driver response from the engine under acceleration," says Herr Gesenhaus. "Like other CVTS, our system increases the engine speed a little bit as the car gains speed, so it behaves more like a conventional manual or aumatic gearbox.'

Driving the Audi A6 mltitronic is a revelation. With the selector lever in D, it behaves just like a conventional automatic, except that it moves off from rest with less of a surge, accelerates more smoothly and never changes gear. Progress is absolutely free from jerks and jolts as the ratio varies constinuously from low through to high. On the autobahn, peak engine revs come up sooner than expected, but then the rev counter needle drops to lower revs at cruising speeds. With the selector in the alternative tiptronic slot to the right of the main gate, the P,RN,D display in the instrument cluster changes to 6,5,4,3,2, 1. In the lower five gears - sixth is an overdrive - the transmission is transformed into a close-ratio sports gearbox. It can he operated by flicking the lever back or forth or more conveniently, clicking on the plus/minus buttons each side of the steering wheel crossbar.

The really clever bit is that you don't really have to use them at all, except to induce engine braking into corners or hold a lower ratio downhill. In sequential mode at full throttle, the transmission changes up automatically as soon as the engine reaches peak revs through each of the gears in turn. And as you slow down, it always prevents the engine stalling by downshifting at the last minute. In between, it hangs on to each step until you reach peak revs or flick the switch. living an uncanny combination of total control and amazing smoothness.

Performance and economy figures prove multitronic efficiency. m 0, 62mph, an A6 2.8 multitronic like 1 drove in Germany beats the five,speed automatic by 1.3sec and the five,speed manual by a tenth. EU figures show it consumes 0.9 litres less fuel per 1 0Okm than the auto. It even betters the manual car's economy by 2 per cent. In the UK, where the front drive A6 2.8 isn't sold, multitronic will be offered in the A6 2.4 in late summer. 'Eventually multitronic will become the only Audi automatic," Herr Gesenhaus predicts. "We are working on new applications for both diesel and quattro models, which should he ready in the next couple of years."
------------

[Marco94]

The so-called "rubber band effect" is indeed the biggest problem for the CVT in general. I have just dug up two articles about the Audi A6 2.8 Multitronic (in the UK it'll be a 2.4) from the CVT in German magazine "Mot." #22 in 1999 and #7 in 2000. I will see if I can scan some pictures, but don't hold you breath.

In the article in Mot 22/1999, the same Reinhard Gesenhaus as in the previous posting, mentions that the shift strategy used by the Multitronic is inferior to a normal CVT shift stratagy: "With employing this stratagy we are mainly taking a psychological problem into consideration. With a regular CVT strategy we can definitly accelerate quicker, but it will sound to the driver as accelerating slower." Only for kick-down, do the use a conventional CVT strategy.

The reasons that the Audi is actually more efficient than the manual gearbox variant is down to the detail engineering they put in the whole system and the fact that they do not use a torque converter, but a wet plate clutch. I know what I want to buy!


Marco.

[Marco94]

Here are the scanned pictures I promised

The complete transmission system:



The shift strategy used:



The working of the CVT:



The chainbelt



Marco.

++
These pictures have been removed due to limited webspace. I'll see if I can get them back again.

[Pine]

Originally posted by mat1
Yes. Where is LJKS now? is he still writing? Or is he retired?

Mat1 [/B]

A friend told me recently that Setright writes of Autocar now.

[MN]

New Nissan Skyline 350GT-8 CVT.
-8 means this CVT has 8-speed manual mode.
In manual mode, you can change it either by the steering wheel paddles or the stick, it only takes 0.2 sec.
I only saw CVT photos with a BELT and two moving pulleys before but not like this one.
Two pulleys(or disks) on engine side and drive shaft side are fixed but little round thingy(Power Roller) between them will tilt to transfer power.


If you understand Japanese...
It has video @ http://www.webcg.net.../000011160.html
Drive impression @ http://www.webcg.net.../000011136.html

[MN]

Forgot to mension.
Nissan engineer in the interview says avarage driver can shift gears in 0.8 sec.
Professional racer in a long destance race who has to look after gears may shift in 0.5 sec.
Professional racer in qualify sessions may shift gears in 0.2 sec.
Driving Skyline 350GT-8 in manual CVT mode, "Anyone can shift as fast as Schumacher in qualify" he says.

[imaginesix]

Originally posted by MN
Forgot to mension.
Nissan engineer in the interview says avarage driver can shift gears in 0.8 sec.
Professional racer in a long destance race who has to look after gears may shift in 0.5 sec.
Professional racer in qualify sessions may shift gears in 0.2 sec.
Driving Skyline 350GT-8 in manual CVT mode, "Anyone can shift as fast as Schumacher in qualify" he says.

Shifting gears in 0.8 sec in a road car is not a good thing unless it is hooked to a very low-inertia, narrow-powerband engine. If the engine takes 1 sec to drop from one rpm speed to the next, then shifting gears in 0.2 sec will cause nothing but damage to the driveline, engine, handling, and driver's ego.

[gil]

It was gratifying to see Marco'94's comment regarding the CVT as being the "only true transmission"! My son and I have been racing an open wheel car in SCCA (Sports Car Club of America) sanctioned events for sometime ( me 20+ yrs, he 6 yrs), its tube framed, fiberglas or carbon fiber bodied, uses a 500cc, 2 cyl.oil injected, 2-36 mm Mikuni carbs,water cooled engine (Rotax, AMW and Kawasaki engines are spec'd by the SCCA) coulpled to CVT drive line, the same as used in snowmobiles. The class is called Formula 500 or f5's, we're usually grouped to race with F-Vees, Club Fords and sometimes with Formula Fords. Pump gas is required and during post race inspections the fuel is checked, also post race weigh-in is 800lbs., less than will get you disqualified. The wheel base max is 80", max width is 55" and OAL max is 150". Also the specs limit the shocks(dampers) to a peice of rubber not to exceed 1" thick and 2" in dia., one per corner, also the spec's call for a solid or tube-one piece rear axle. Spec tire size is 19.5x7.50-10 on the rear and 19.4x6.50-10 on the front- numerous compounds, in slick and rain treads are available from Goodyear and Hoosier.
As to performance, a 0-60 mph run is about 4 secs., our top speed has been clocked at Road Atlanta at 144 mph, at our "home" track- Summit Point Raceway in West Virginia, the lap record is 1.17.4(not by us!). Summit Point is a 10 turn-2 mile long, fairly smooth track, sometimes used as a test track for some CART teams, the overall track record for SCCA events is 1.10 - in a Formula Atlantic, the CART guys do 1.00's!
The f5's are entry level or for the budget minded (me !- we can campaign 2 f5's for the cost of one FF or FC), so having pleasing engine sounds will be something the spectators and f5 drivers will have to be patient for,either the next race or the next class up !
As mentioned by Ray Bell, the CVT lacks the challenge of a gear box and the "heel and toe" one can practice when decelerating. The f5's with the CVT's don't offer the opportunity to capture the
engines compression when braking, you use left foot braking and right foot "go". Also the "foot box" is small and slapping ones foot back and forth is not feasible, so to be competitive in this class you have to use left foot braking and alot of "trail brake" (power-on braking).
The engines power band comes on at 5200 rpms thru 8100 rpms, a competitive driver when braking will know how much to let off the throttle to keep the clutches from going into "rest", minimizing the "spool up" time, a must for good lap times! On several corners at Summit Point the driver won't lift his foot but does tap the brake going in to allow lateral transfer- if the sidewalls aren't working you're headed for the gravel trap!
Indy Ian's clutch remarks were on target, for many tuning f5 CVT's is almost a "black art" !
You have 2 clutches- the primary/mounted on the engine and the secondary/mounted on a jackshaft(which has mounted on its opposite end a sprocket-if chain drive or a cog gear for a Gates belt which powers the rear axle). Each of these units are "tuneable", in the primary different weights and their profiles can be used to control engagement rpms as well as using a different tensioned main spring. The secondary also has a spring that can be replaced with a different tensioned spring and also a helix or sometimes called a "ramp" that can be changed to control "backshifting", the amount of tension is again adjustable even when another spring is used. If ones secondary is not adjusted properly you go into a higher ratio before you're into the power band- life and everything will pass you by- its like throwing out the proverbial anchor or you could be going flat out and doing 60 mph!
The f5 class has a web site you may want to check out- f500@www.f500.org - plenty of technical info, cars and parts for sale and has some links to manufacturers and component suppliers.
Gil

[lateralforce]

Damnit. I don't know who to believe now. I had the opportunity recently as an exchange student to join a school visit to ZF (Zahnfabrik) in Friedriechshafen (hope i spell it right) and they make transmissions, steering systems and a few other things for Audi, Porsche and such. Now in a presentation by a company representative revealed that they do make the CVT, however they say in terms of efficiency and performance CVT isn't as different compared to their latest 6 speed auto transmission. The guy also added the future of CVT will solely be depending on how the market trend is going..if a lot of carmakers start to have faith in CVT then it will gradually phase out the current auto transmission..... So who is right now?

[da Silva]

This is a little OT but you guys could probably have an good answer to this one:
The quote is from Max Mosely

Question: I thought that there was controversy going on about gearbox systems?

Max Mosley: There’s a little bit of controversy about twin clutch gearboxes. Cars these days have seven forward gears, and obviously the more forward gears you are allowed, the narrower the torque band for the engine can be and therefore the greater the power – you can have a really peaky engine – particularly now it’s all done by computer. You couldn’t have a seven speed box if you had to do it manually. But they can’t have more than seven speeds so if you then narrow this right down, what you really need is CVT. Now that’s illegal, but maybe if you have seven gears… with these twin plates clutches, what happens is that one set of gears is engaged while the other one is driving the car, so the gearchange becomes almost instant. You just swap clutches, instead of having to engage different gears. Now you could arrange it that the clutches worked in such a way that on that particularly awkward corner where fifth was too short and sixth was too long, it just got you over that little bit. It would generate a bit of heat and so on, but it would just get you over that bit. I’ve explained that very badly and crudely, but you can see the essence of it. Well, we make it clear that you can’t do that and of course we will be looking at the software to see that it doesn’t do it. But that’s in essence, as I understand it, the danger with the twin clutch system. I think there is more than one team that have these and they’ve existed for a long time. If they are just used as a means of speeding up the gearchange, it is unobjectionable. It’s only if it is used to expand the range of the gearbox.

I rememberd Mika Häkkinens car failed on the final lap od the Spanish GP 2001 and they
said it was the clutch that failed. Could someone please explain this technic with the double clutches?

[AdamLarnachJr]

A blurb on CVT's, Ford is testing a new truck model which is designed to be used by their 7.3 litre Powerstroke motor, which is a 550ft. lbs turbo diesel.

"We started out three years ago looking for a way to improve our truck and SUV fuel economy, and were totally unprepared for how spectacular the results would be from a driver's perspective. What we've ended up with is a tranmission which offers refinement, smoothness, and responsiveness unmatched anywhere in the truck or SUV industry." -Vance Zanardelli, manager, Transmission and Engine Systems

Continuously variable transmissions hold potential for significant fuel savings, however up until now, their use has been limited to relatively small vehicles with engines generally 3.0 liters or less.

After three years of development, Ford Motor Company engineers are testing a more robust CVT design in a Ford Expedition sport utility, with its standard 215 hp, 4.6-liter V-8 engine. This represents the first time this technology has been applied to a full-size truck.

The most common CVT technology uses a belt-and-pulley system to transfer torque through a range of drive ratios. These systems are typically used in front-wheel-drive applications, with relatively small engine sizes, due to an inherent limitation in the design.

As the belt-and-pulley components are made larger to handle the higher torque demands of more powerful engines, the belt eventually gets so large that it generates excessive centrifugal force as it moves around the pulleys. The clamping force required to hold the belt in place eventually generates unacceptable mechanical losses.

The research system uses a Traction Drive CVT design that offers several advantages. In this architecture, called a "variator," two discs are arranged on either side of a tilting or articulating roller - one disc for torque input, the other for output. The roller angle between the discs determines the final drive ratio. There is no belt.

While torque in the belt-and-pulley system is transferred via metal-to-metal friction between the components, the Traction Drive CVT establishes a fluid film between the polished surfaces of the roller and discs. This boundary layer - about one micron thick - develops enough shear strength to transfer torque between the components, without actual metal-to-metal contact. These qualities allow it to be upsized for larger uses, as in full-size trucks or sport utilities.

Ford Motor Company engineers estimate it could handle engines up to 7.3 liters in displacement.

The fluid is completely synthetic, developed specifically for this purpose. Where conventional transmission fluids would simply allow the smooth bearing surfaces to slip past each other, the new fluid - dubbed "liquid gear teeth" by engineers - begins to act more like a solid under the high pressures generated in the Traction Drive CVT, and adheres to the molecules of the metal surfaces. The roller and discs can exert up to 600,000 pounds per square inch against each other.

The Traction Drive CVT technology - for which Ford and Mazda hold 10 patents - holds out the promise of considerable fuel economy improvements for these vehicles, in addition to the benefits of smoothness and refinement that come with eliminating the relatively large steps between gears of a conventional automatic transmission.

For example, Ford Motor Company engineers are projecting a potential, immediate 10-percent fuel-economy improvement in the test Expedition, with only calibration changes in the powertrain and without sacrificing either engine power or emissions performance.

In fact, project engineers expect both overall emissions and vehicle drivability to improve with the Traction Drive CVT.

One reason for this immediate fuel economy gain is the wide gear ratio that the CVT offers. The Expedition's four-speed automatic transmission operates over a ratio of approximately 4:1 from low gear to overdrive. The Traction Drive CVT fitted to the test vehicle uses a 7:1 ratio range.

In highway cruising and most daily operation, this allows a much lower engine speed - perhaps 500 revolutions per minute less. When more power is needed, such as for passing or climbing a steep hill, the CVT seamlessly steps down, engine speed rises, and more power is on tap almost immediately.

It reacts so quickly to engine load changes that it can shift from lowest to highest gear within about six engine revolutions - far quicker than would be necessary in the real-world environment.

In another economy benefit compared with traditional automatic transmissions, the Traction Drive CVT design allows a very aggressive lockup strategy at the torque converter, reducing mechanical loss.

Ford engineers are currently refining and calibrating the newly completed test vehicle, as they begin field trials to confirm the fuel economy benefits.

CVT Joint Venture

Ford Motor Company also is pressing ahead with plans to develop belt-type Ecotronic variable transmissions to bring the benefits of CVT wide gear ratio span to front wheel drive vehicles, such as passenger cars and small sport utilities. The new Econtronic family of transmissions can handle up to 174 foot-pounds of torque, greatly expanding the possibilities for belt-type CVTs. "

[desmo]

As seems to be more and more often the case, road car technologies far outstrip those now seen in F1 due to the ridiculously narrow technical regulations that are currently in force. F1 will apparently continue to fall further and further behind the technology seen in everyday roadcars. Maybe they're going for some sort of retro-chic appeal.

[MN]

Originally posted by imaginesix

Shifting gears in 0.8 sec in a road car is not a good thing unless it is hooked to a very low-inertia, narrow-powerband engine. If the engine takes 1 sec to drop from one rpm speed to the next, then shifting gears in 0.2 sec will cause nothing but damage to the driveline, engine, handling, and driver's ego.

Tell Nissan about it, they may not know that.;)
Skyline 350GT-8's engine is 3.5-liter DOHC V6, 272hp/6000rpm, 36.0kgm/4800rpm.

[Superliner II]

Was there such a thing as a Williams F1 car in the eighties with a constantly variable transmission?

I dont recall reading or hearing about one. I'm just trying to figure out if i am being fed BS by someone who calls himself a motorsport engineer ;)

[Wuzak]

There was a Williams-Renault fitted with a Van Doorne (?) Continuously Variable Transmission in the '90s, undergoing tests.

I can't recall if CVT was banned in 1994, or if this was a later development.

The system was showing some good promise at the time of its banning.

[gshevlin]

Yes, Williams built a CVT and tested it for at least 2 years. Mark Blundell carried out the majority of the testing when he was the Williams test driver. However, the FIA banned CVT in 1992, rendering the Williams research null and void. This was not the first time the FIA had done this to Williams - they banned cars with more than 4 wheels at the end of 1982, ruining Williams' plans to run their six-wheel car in the 1983 season.

[desmo]

There are several threads archived here on the CVT, I've merged this one with the most recent.

Here's a bit from Peter Wright's book "F1 Technology" on the topic:

"Step-less gear ratio changing permits the engine to operate within a narrow rpm band, or even at a discrete rpm, instead of requiring an operating band of appoximately 2000 rpm. The engine designer can then tune the engine for that rpm, avoiding any compromises in intake and exhaust lengths or valve timing. The increase in available poiwer (the integral of power throughout a cycle of accelerating through the speed range) must outweigh any loss in efficiency or increase in weight.

Williams started research into CVTs in the late 1970s, initally with a system based on the Perbury system. At the time, Williams was developing the Cosworth DFV and was seeking ways to achieve a power advantage over other DFV users. The Perbury proved to be too heavy and was encumbered with IPR problems. Some years previously, Van Doorne had raced an F3 version of its rubber belt and cone CVT, based on a system sold in DAF road cars. It was reasonably competitive but unable to take the power of an F1 engine. Williams decided to work with Van Doorne to develop a steel belt version, suitable for F1. The issues of engine starting, engaging drive, reverse, and the hydraulic and electronic control integration with the engine all had to be solved. The gearbox that was mounted in a Williams FW15, powered by a specially tuned Renault 3.5 liter engine, weighed 6 kg more than the six-speed one it replaced and was approximately 93% efficient, compared to 96% for the spur gear variant. This appeared in testing in 1993 but was banned as part of a package of control system regulations, at the end of the year, before its performance advantages could be demonstrated. History does not relate whether the ban was a result of the fear of having to develop competing systems, when Williams had been working on the principle for so long, or the thought of 24 F1 cars, all with their engines continuously at peak rpm. CVTs are now becoming accepted road transmissions, competing with automatics and a variety of computer-controlled gear change systems. Audi launched a steel and cone-based CVT system in 1999, capable of accepting 280 Nm torque, which endows its A6 with better performance and economy than a manual five-speed transmission, equivalent version. F1 could have promoted and hastened the development of what may become the transmission of the future."

[Wuzak]

Originally posted by gshevlin
This was not the first time the FIA had done this to Williams - they banned cars with more than 4 wheels at the end of 1982, ruining Williams' plans to run their six-wheel car in the 1983 season.

How many of those 4 rear wheels were driven?

[desmo]

All four rears were driven in the '81 FW07D and '82 FW08C (be advised, the taxonomy is controversial) as was the earlier March with four rears. The advantage was said to be primarily aero, although anecdotal reports were that the Williams both did standing starts very smartly and was notably steerable under hard throttle. Neither made it to a race as far as I know.

[Wuzak]

Originally posted by desmo
All four rears were driven in the '81 FW07D and '82 FW08C (be advised, the taxonomy is controversial) as was the earlier March with four rears. The advantage was said to be primarily aero, although anecdotal reports were that the Williams both did standing starts very smartly and was notably steerable under hard throttle. Neither made it to a race as far as I know.

Surely that would have made them illegal BEFORE they were planned to be raced - 4wd had been banned for many years by that stage.....................

[Wuzak]

I had heard that test showed that it would be very good in wet conditions.

They could run wets on the front pair of driven wheels, and, becuase they were close behind, dries on the rear set of wheels, giving great traction!