autogyro's Transmission Concept

[autogyro]

It seems that F1 gearboxes do not clever mechanical systems to provide a seamless shift. F1 double clutch geabroxes are not banned per sé, but the team does need to prove that the 7 different ratios are engages individually, as a double clutch set up could be phased\slipped to provide all but a CVT like solution. McLarens DC was tried on the MP4-19, while Williams raced a DC in 2006 with horrendous reliability.
Nowadays the teams run a conventional 7 speed dog gearbrox, but with two seelctor drums, each one looking after alternate gears. thus one selector is engaging the next gear, while the other disengages the current gear. With a throttle lift or dab of clutch, the electronic\hydraulics phase the two movements into a seamless selection. Seems a simple solution compared to the mechanical mayhem of Weisman or Zeroshift.

Yep, but still a torque sapping layshaft box with at least four gears under engagement load in each gear and all the rest being powered round in an oil bath.
Pointlessly.

[autogyro]

It seems that F1 gearboxes do not clever mechanical systems to provide a seamless shift. F1 double clutch geabroxes are not banned per sé, but the team does need to prove that the 7 different ratios are engages individually, as a double clutch set up could be phased\slipped to provide all but a CVT like solution. McLarens DC was tried on the MP4-19, while Williams raced a DC in 2006 with horrendous reliability.
Nowadays the teams run a conventional 7 speed dog gearbrox, but with two seelctor drums, each one looking after alternate gears. thus one selector is engaging the next gear, while the other disengages the current gear. With a throttle lift or dab of clutch, the electronic\hydraulics phase the two movements into a seamless selection. Seems a simple solution compared to the mechanical mayhem of Weisman or Zeroshift.

Yep, but still a torque sapping layshaft box with at least four gears under engagement load in each gear and all the rest being powered round in an oil bath.
Pointlessly.

CVT'S AND TVT's for high torque transfer need to use a lot of power to operate the movement of cones (Van Doorne, or toloroidal disks (torotrak). In the torotrak high power is needed to effect the operating condition of high pressure needed to maintain a hydraulic friction drive from the special fluid between the disks. It also needs a lot of cooling. Both use too much power to operate and are very driver unfriendly. Like driving a banana in a bowl of custard. I told Patrick this under a CA at the time. However flywheel energy storage is a good direction for Kers but without the torque sapping torotrak connection. CVT's and TVT's are useful in instrumentation and light use. Flywheel's work in buses etc but will only work in F1 under more open regulations for Kers.
My system has constantly variable magnetic shift overlap, that actually adds to the torque transfered but still maintains seven or any number of fixed ratios.
These ratios can be selected sequentially, in any order or any number together, depending on the control system parameters. Don't forget, no clutch either.
Technical innit!
Oh and it is not perpetual motion you silly boys.

[F1_eng]

As far as I can recall, this has not come accross my desk and I have what people would reagrd as a pretty high engineering job in a formula 1 company.

Any details of a patent or test/concept results?

[autogyro]

As far as I can recall, this has not come accross my desk and I have what people would reagrd as a pretty high engineering job in a formula 1 company.

Any details of a patent or test/concept results?

Andy Scott at Hewlands has seen a schematic ESERU under confidentiality, if that helps.

[riff_raff]

autogyro:

Regarding some of the comments in your previous posts:

"In support of current layshaft gearboxes. Development in road vehicles and race cars has always mainly been aimed at the engine and the chassis/aerodynamics."

This is because a well designed manual transmission (MT) is already a very efficient device, so there's not much improvement to be gained here. A good spur gear automotive transmission is likely better than 98% efficient, input to output.

"The IC engine is almost at the end of any efficiency development and the essential improvments have to be found in other areas. The transmission is one of these prime areas of development."

The recip piston IC engine is improving in thermal efficiency and exhaust emmissions at a much faster rate than any other current type of automotive propulsion system, including battery-electric or fuel cell. And this has been the case for the past 20 or 30 years. Current production automotive diesels are now approaching BTE rates of 40% or better, which is an improvement of at least 35% over just the past decade. And newer combustion cycles, like HCCI and PCCI, will result in even greater thermal efficiency gains for the IC engine when they are employed.

As for transmissions, the only type of transmission that is getting more efficient is the conventional automatic (AT). But these efficiency gains come at the cost of much greater complexity, due to the greater number of gear ratios they now employ. In fact, a modern 6 or 7 speed AT is easily the most expensive and complex part of the car it is installed in. The basic manual transmissions has not gotten any more efficient over the last 20 years. However, dual clutch automated manual transmissions (AMT's) with sophisticated digital controls, are now displacing the conventional automatics, due to their lower costs and better efficiencies.

"A good gearbox can also improve the efficiency of a pure electric vehicle."

The reason this is true is due to the inherent operating limits of the typical AC motor used in automotive electrical propulsion systems. Many auto OEM's (such as GM) now use a "dual mode" (ie. dual motor) electrical drive system architecture in their hybrid car transmissions, just to address this very problem.

By the way, I'd be very interested in seeing your transmission design. Where can I get some more details?

Regards,
Terry

[autogyro]

autogyro:

Regarding some of the comments in your previous posts:

"In support of current layshaft gearboxes. Development in road vehicles and race cars has always mainly been aimed at the engine and the chassis/aerodynamics."

This is because a well designed manual transmission (MT) is already a very efficient device, so there's not much improvement to be gained here. A good spur gear automotive transmission is likely better than 98% efficient, input to output.

"The IC engine is almost at the end of any efficiency development and the essential improvments have to be found in other areas. The transmission is one of these prime areas of development."



The recip piston IC engine is improving in thermal efficiency and exhaust emmissions at a much faster rate than any other current type of automotive propulsion system, including battery-electric or fuel cell. And this has been the case for the past 20 or 30 years. Current production automotive diesels are now approaching BTE rates of 40% or better, which is an improvement of at least 35% over just the past decade. And newer combustion cycles, like HCCI and PCCI, will result in even greater thermal efficiency gains for the IC engine when they are employed.

As for transmissions, the only type of transmission that is getting more efficient is the conventional automatic (AT). But these efficiency gains come at the cost of much greater complexity, due to the greater number of gear ratios they now employ. In fact, a modern 6 or 7 speed AT is easily the most expensive and complex part of the car it is installed in. The basic manual transmissions has not gotten any more efficient over the last 20 years. However, dual clutch automated manual transmissions (AMT's) with sophisticated digital controls, are now displacing the conventional automatics, due to their lower costs and better efficiencies.

"A good gearbox can also improve the efficiency of a pure electric vehicle."

The reason this is true is due to the inherent operating limits of the typical AC motor used in automotive electrical propulsion systems. Many auto OEM's (such as GM) now use a "dual mode" (ie. dual motor) electrical drive system architecture in their hybrid car transmissions, just to address this very problem.

By the way, I'd be very interested in seeing your transmission design. Where can I get some more details?

Regards,
Terry

Good informative post terry. I have to disagree with your efficiency figures however. The torque in to torque out figures are much more complex and depend on many variables. One major factor being the input RPM another the level of input torque. The effects on IC engine efficiency has also to be calculated plus the effect on electric power input in a hybrid or pure electric.
The 98% given for some modern automatic multi clutch spur gear, lock up converter transmissions and layshaft boxes is the best condition figure and a marketing con. Usually it is calculated with converter locked and the least relative rpm between components and lowest o/s revs in that condition. It is not a fair definition of torque efficiency in over all use. Neither is it for the layshaft gearbox. For one thing the loss of torque through oil windage increases markedly with RPM. Of course any dual shaft box must have twice the laygear losses!
(years ago I used to machine off gears in oval circuit gearboxes to reduce oil windage and I never came close to 98% torque efficiency)

My system has no gears running in an oil bath, there is just an oil mist for tooth lubrication and some cooling, the actual gears have sealed bearings. Only one gear set is engaged for each ratio and the others do not rotate at all unless engaged. In direct top gear there are NO gears turning relative to each other and only one gearset/shaft support bearing. So each gear use in my system would give half at least of the torque loss in all conventional gearboxes, (over all in all gears) and as near to zero torque loss in top gear as you can get (one bearing and far better).

IC engines at 40% are closer to the inefficiencies of external combustion steam engines than to electric motors. It is the currently available gear systems that are defeating major improvements in electric vehicles. The conventional systems do not like the instant and high application of torque given by electric motors. This is why the Tesla went back to one direct gear (the three speed unit they tried first would only last 2000 miles). Which is not as some people say 'no gearbox' because there is a ratio drop. It is a one speed gearbox. Most small EV use belt drive for cost/efficiency reasons, they are yet to realize the horrible noise result in practice. Unfortunately the transmission engineering has not yet been fully appreciated by the highly qualified electronics engineers who by and large have very little experience of practical vehicle development and even less in transmission experience. It annoys me that they use the superb power delivery of electric motors to try and convince other engineers with years of experience in vehicles that no gearbox is needed.

All this is why it was a criminal move by Fota, when it forced Kers to be dropped.
Kers is an area of development that is essential for F1 and the continued development of vehicles through the EV revolution. The ban will hold back road vehicles by years and allow inefficient first generation EV's into production to bolster the big car companies through the government hand outs they now live off.

My ESERU can also be configured as an 'integral' part of a far more efficient energy harvesting unit for either Kers or non Kers use. It can also be used for a multi hybrid system or a full electric application for road use.

[F1_eng]

What are you trying to say?

I dont understand where you are going with this, what type of gears are you going to be using in this system?
What is the packaging relative to an equivalent current production car gearbox?

One bearing support sounds a little strange to me but again, without a description of the transmission, I can't discuss any further.

I agree that in some cases, transmission efficiency figures can be a little missleading but what can you do. We design gearboxes to be out of optimal alignment during low torque transfers purely so we have optimal during a higher percentage of the opperating range. Normal transmission testing would not convey this fact and the gearbox would usually end up with a poor transmission efficiency. Road cars have different requirements again so its very difficult to compare.

[autogyro]

What are you trying to say?

I dont understand where you are going with this, what type of gears are you going to be using in this system?
What is the packaging relative to an equivalent current production car gearbox?

One bearing support sounds a little strange to me but again, without a description of the transmission, I can't discuss any further.

I agree that in some cases, transmission efficiency figures can be a little missleading but what can you do. We design gearboxes to be out of optimal alignment during low torque transfers purely so we have optimal during a higher percentage of the opperating range. Normal transmission testing would not convey this fact and the gearbox would usually end up with a poor transmission efficiency. Road cars have different requirements again so its very difficult to compare.

Gears in the ESERU are simple noise limiting minimal side load helical.
The FI ESERU geartrain goes into a 110mm cube. It is also an electric generator and an electric motor all in one unit. The gearsets in it weigh less than the current layshaft gears in a layshaft dog engagement F1 box.
The input of the unit splines directly into the engine crankshaft for support and turns with it and the output is supported in one bearing.
Because there is no clutch 110mm is the length from the crankshaft to the output shaft F1. Useful? Size for road vehicles would be larger but still much smaller than any current gearboxes in use.

As far as I know there are few companies making gearboxes, packaging geartrains or developing cars who have full test facilities for transmissions these days. Most have nothing at all and simply follow long established practice using geartrain principles from two centuries ago and the unchanged figures from books dialed into CAD programs. I know this from personal experience and from my time developing and using dynamometers for engine and transmission testing both static and mobile. Derek Gardener did have a reasonable test cell for automatics at Letchworth when he was running Borg Warners years ago. I used it a few times. Sadly long gone with most of our superior British engineering.
Testing automated selection systems is just playing with the external bells and whistles, not the geartrain concept itself.
The effect the gearbox has on the efficient transfer of torque from the engine to the drive axle is simply accepted from the quoted manufacturers figures these days and all they look at, are the ratios to choose, the weight, bulk and the burst loads. It is only the complete power train that is measured at engine input and output to axle. They can even play tunes with the Unit control system on the exhaust note but almost completely ignore the gearbox, other than for somewhere to connect the suspension etc.

I am sorry I cannot post you a description of the ESURU but I am semi retired and am not prepared to either give the idea away (as I have on a number of other occasions with other transmission ideas) or spend huge effort and cost traveling the world selling anymore. If anyone wants to know more they will have to pay.

[F1_eng]

I can't comment on other companies but we have everything we need to fully test and benchmark our fully in house designed and manufactured carbon gearbox.

Another company which does a lot of transmission/driveline testing and are a very good and interesting company is GKN.

I did quite a lot of work on moving bearings within a gearbox a while ago, trying to change the whole structure of the gearbox. Some ideas we are using and some I came up with were cetrainly not going to work within the rules. One interesting concept i adopted was using another gear that rotates in the same way as output to house a bearing internally, that meant bearing forces were reduced and also the dynamic bearing loads due to rotational speeds were considerably reduced since the speed of the gear that housed the bearing was reducing the overall bearing speed.
As with most things, this intorduced a possible lubrication issue which meant some gears had to be re-designed to allow oil flow.

[autogyro]

I can't comment on other companies but we have everything we need to fully test and benchmark our fully in house designed and manufactured carbon gearbox.

Another company which does a lot of transmission/driveline testing and are a very good and interesting company is GKN.

I did quite a lot of work on moving bearings within a gearbox a while ago, trying to change the whole structure of the gearbox. Some ideas we are using and some I came up with were cetrainly not going to work within the rules. One interesting concept i adopted was using another gear that rotates in the same way as output to house a bearing internally, that meant bearing forces were reduced and also the dynamic bearing loads due to rotational speeds were considerably reduced since the speed of the gear that housed the bearing was reducing the overall bearing speed.
As with most things, this intorduced a possible lubrication issue which meant some gears had to be re-designed to allow oil flow.

Unfortunately in a conventional layshaft F1 gearbox the geartrain forces are off center, which requires a need to design a casing with complex force calculations.
I thought most cases were now metal partly because of this and the limitations of carbon fiber? I see the thinking behind your bearing idea (useful) but you are still compromised by off set burst loads. Oil flow needs and immersed gears because of this forced requirement.It will always result in high oil drag (windage, and fatigue/stress fracture potential in drilled components.
My ESERU has balanced burst loads (I am going no further here) and the casing can be a much better design. Each gear has a sealed bearing with even radial load and no oil feed requirement.Shaft support bearing (the only one that rotates all the time)and shaft needle bearings which do not are oil fed. Mist is released from them for tooth lube. Oil is dry scavenged. Cooling is another method, which I will not mention.
I am interested in the test kit you use to test the gearbox. Do you simulate all input variables and measure output in all shift, drive and over run conditions?
It would not all be necessary with a known geartrain layout, if as you previously said, you are only interested in a narrow operating range. Most if not all of the data you need should be available using the computer and past data input. Not a method I am fully in agreement with but should be good enough for what you want.
Not a criticism in any way but there is more than enough layshaft data available to package an F1 conventional gearbox without test rigs.

GKN did some special work on CV joints and drive shafts for me years ago for a record Sahara crossing I did with Ginger Baker. They also supplied strengthened drive shafts for an automatic saloon I built in the 70,s. Good to see they are still up there.
Pity about AP, they still do clutches I hear but not the automatic gearboxes I used to help with under contract to Leyland special tuning. I once did a racing AP1 four speed auto that won a couple of races back then.

[autogyro]

So fi-eng, are you using Kers next weekend?

[autogyro]

So fi-eng, are you using Kers next weekend?

Having thought of your support bearing in gear idea f1-eng, surely the bearing loads reduced on the primary support would be transfered to the mounting gear and increase the tangential load on the secondary gear support bearing? I agree that primary bearing rpm would be reduced, so perhaps there is a net gain. Depends how bad the original casing design was.
I suppose your aim was to reduce and even out the bearing support positions at the casing ends to reduce off set loads and give the casing less work to do.

Because my ESERU only has one support bearing, I have even considered a skeleton casing to allow air cooling through the casing and aero improvements. A turbine blade design for the last internal component could be used to force air for cooling and also increase airflow over the rear diffuser. The operating components needed in the casing could be designed to allow air flow through them.

[Jersey Tom]

So, just to clarify, your proposed unit which takes the place of the clutch and multi-speed gearbox, fits in a 4x4x4" cube?

[autogyro]

So, just to clarify, your proposed unit which takes the place of the clutch and multi-speed gearbox, fits in a 4x4x4" cube?

110mm yep, I just measured the drawing again.

[autogyro]

So, just to clarify, your proposed unit which takes the place of the clutch and multi-speed gearbox, fits in a 4x4x4" cube?

110mm yep, I just measured the drawing again.

Oh I forgot to add, it is also a major part of either a Kers system
(new generation), a hybrid geartrain, or a full electric vehicle power train.
Yep all in 110 mm.