Yes, I just had to do the best I could with the figures available. Older automatics are worse I believe and newer ones are getting better. I have figures for an automatic (with friction launch clutch) being always greater than 92% efficient from the Frankfurt motorshow so a manual couldn't beat that by more than 8% in any normal driving circumstances. (There were some exceptions to the efficiency, like high RPM, so the manual could still be a reasonable amout better at top speed.) It has power through two compounded planetary gearset, so efficiency will sometimes be of the order of 96%*96=92% so there is still room for improvement in planetary automatic design.
I have seen designs where only one planetary gearset is charged with power at a time. With a geared input and output shaft having losses of say 1% each, losses would tend to be around 1% (input) + 1% (output) + 1% (oil pump) + 4% (planetary gear) = 7% when in most of the gears, and with a couple of direct drive modes for the two more commonly used gear ratios, so about 1+1+1=3% losses when pottering around.
So let's say 95% average efficiency is round-about a theoretical maximum for planetary automatics in design, and since things never go as well as expected in practice, I'd say 94% average is maybe the best we will see, even if greater than 92%, ie 92% minimum (with caveats) is already being quoted by planetary automatic transmission designers.
For the Europeans, based on countershaft designs rather than planetary designs, due to their history of manuals instead of automatics, they can improve further. In my book, any drive-by-wire transmission has blurred the definition boundaries of the words automatic and manual. The transmission does what the software tells it, and that can be based on driver input, eg +/- controls, or software decisions: we are a bit fast so change up.
The VW dual clutch transmissions are effectively like two transmissions, one for A={1,3,5} and one for B={2,4,6}. A clutch is release for the offgoing gear A or B, and a clutch activated for the oncoming gear B or A. Mechanical efficiency of these transmissions should pretty much equal those of a manual. There is a problem that only odd-to-even or even-to-odd gear changes are possible, so for example 6 to 2 would have to be something like 6 to 5 to 2.
There are Austrailian designs where the dual clutch transmission has been taken a step further and each of the {1,3,5} and {2,4,6} transmissions has powershift capabilities. In that case, 6 to 2 would be an internal powershift and 6 to 3 would work like a normal dual clutch gearchange. So all the creature comforts a six-speed automatic should be available in the future with manual gearbox type mechanical efficiency.
Statistically efficiency (as in CO2 g/km) will go up due to better (computer controlled) gear selection and the powershifting.
[The European method of testing efficiency, grams of CO2 per kilometer, is superior to the American system of measuring miles per gallon. If America sticks to miles per gallon then there will be pressure to switch to 'more efficient' diesels, which get more miles per gallon. However, it takes more oil out of the ground to make a gallon of diesel than to make a gallon of petrol. Secondly, petrol is mainly Carbon 6 and Carbon 8 molecules boiling below 180°C whereas diesel is mainly C 12 and above molecules boiling 160-400°C. Due to the bigger molecules, diesel has a higher energy density, and so it darn well ought to give more miles per gallon. End result: measure CO2 and not MPG, to have a fair test! Last thing to say about diesels is they have higher mpg because of the higher compression ratios too. With petrol direct injection that difference should be balanced out a bit in the course of time. People have been looking at HCCI engines, homogenous charge compression ignition. That doesn't seem worth the effort, as they cannot precisely predict the moment of autoignition: the solution in my opinion is to take it as close as they can to autoignition and then spark it - viola, high compression ratio and no timing problems.]
Hybrids are another subject. Its really about energy recovery rather than mechanical efficiency, but there is the stealth engine downsizing aspect too which will be significant for countries with oversized engines. For overtaking with an SUV the motor power is there; towing something heavy for a long time at speed, and the motor power won't be there. So stealth engine downsizing by counting motor-generator horsepower in with engine power is good for reducing oil demand and fashion victim SUVs buyers dererve it, but at the same time, there is the potential 'con' for people that actually think they will be buying a work-horse.
Geared CVTs like the Prius?
At the moment, running motor/GENERATOR 1 to power MOTOR/generator 2 is clearly not efficient. Geared (Power split) CVTs have merit, but not as currently implemented.
Perhaps a minor point, but in the interest of accuracy:
'The CVT presumably has an identical gear ratio available to the manual, so the lack of top speed is due to the CVT's inefficiency'
Not necessarily. From the cars.com review:
"While a traditional CVT basically is a belt that changes shape based on the driving situation to provide an infinite number of gears, BMW modified its system. With the Mini CVT, six gears have been programmed into the system so you feel six shifts. You feel no gear changes with a traditional CVT."
Next year's Audi A6 reportedly will have a CVT option that also includes selectable ratios for manual emulation. It seems the OEMs think consumer resistance to this different technology may run pretty deep. I am not positive, but I think the Nissan Murano is programmed as a 'traditional' CVT, and it seems to be selling well in my area. The new Ford 500 could turn out to be the CVT's 'killer app.'
The Mini has different driving modes, so while it can emulate a manual, I think they have also given it a sport mode where it can hold maximum power RPM. (You can read a lot of informed opinions on
- forum - if I remember the weblink.) Presumably the cars top speed would be quoted in CVT mode rather than stepped emulation mode if the speed differed, but you are right to point out stepped gear differences in the calculations.
Its really the same point that was made earlier about possibly having different ratios in the automatics and manuals.
Generally I think manufacturers have been going for top gears which give the highest top speeds on many models - economy models obviously being one of probably many exceptions.
The Nissan Murano - I agree. I watched a program with a desert rally on TV once. Dubai I think it was. I remember the commentator saying something about knowing the Nissan was coming because the only change in pitch of the note was the Doppler SHIFT as it went past. Odd the company with the slogan "shift expectations" is the one doing the work with production CVTs!
The new Ford 500 could turn out to be the CVT's 'killer app'? Its the CFT30 like the Freestyle and Mercury Montego. These are from the troubled Batavia project that ZF just paid €170 million (in the sense that they declared a one-off exceptional loss on their 2003 accounts) to get rid of their half to Ford, after reported investing about $700 million dollars in North America over the past few years, most of that being in Batavia before a single CVT was produced.
In January of last year the Ford Freestyle was reported a CVT vehicle. By July it was being reported as being CVT and 6-speed. It is now being reported as being a 6-speed with an optional CVT. I remember reading an interview with Dave Szczupak and questions about why it wasn't available with the 3.5 litre V6 engine. The Batavia problems seem to coincide with Ford signing with GM to gets its hands on GM's X22F 6-speeds. Predicted volumes for the Batavia CVTs, presumably incuding the Ford Focus CMax minivan with the CFT23, have been steadily dropping from 1 million down to recent predictions of 250,000.
Ford's killer application with CVT might not come from its work with the German ZF, but from the Japanese Aisin, and the Toyota Hybrid System 2 (THS-II).
GM have an AHS-II planned. I don't know if its just a similar name or whether they have plans to use some Aisin hybrids too.
Greg,
Sorry for not replying sooner. My point about TC's being more efficient during launch is that with the torque multiplication, the is more energy driving the rear wheels compared to a slipping launch clutch. If you plot efficiency vs. slip speed, the converter is a curve where a clutch is linear. The converter will rise faster than a clutch, peaking near 0.6 to 0.7 slip ratio before falling back toward the linear clutch efficiency line, intersecting near 0.9 slip ratio. Anytime operating at a slip ratio higher than 0.9, the TC is more energy efficient.
There is something I don't understand here. I don't fully understand what shanba is saying, as I don't know which way round slip ratio works. Naïvely, I would assume 1.0 slip ratio meant 100% slip, ie open clutch and zero slip meant clutch engaged. But that is detail and not the source of my problem.
I take shanba's information on trust. I have no expertise on the subject, but no reason to doubt the information is correct, etc.
So why the better launch performance of manuals with clutches (as per 0-60 MPH figures) if the torque converter has higher mechanical efficiency for most given slip rates (and nowadays can be locked up in that area where its not too good?) ?
Is it because the slip rate can be controlled with a clutch, the driver engages the clutch at a rate he feels is in some way optimum, whereas the torque converter being mechanical presumably engages at whatever rate it sees fit?
That doesn't quite seem to explain it to me, as an quickly fully engaged clutch would result in low RPM and low power, whereas a still-slipping torque convertor might have the engine running faster and producing more power.
And why the nickname 'power sapping device' for something more mechanically efficient than a clutch 90% of the time?
All I can think of at the moment is that a clutch is engaged faster, giving near 100% efficiency at a point in time where the more efficient torque convertor is still slipping at a speed where it is more efficient than the clutch would be, if the clutch were still slipping that much too.
If a computer controlled clutch were programmed to engage at the same rate as the torque convertor in an otherwise identical car, I presume shanba is saying that the torque convertor vehicle would be ahead, (due to higher TC efficiency) correct?
Despite the torque convertors efficiency superiority, its loses are still so great that an engine producing less power due to lower RPM with a closed clutch (nearly 100% efficient) will still out-accelerate the slipping torque convertor noticably.
For drag racing, modulating a clutch correctly to handle the huge power but avoiding wheel-spin is too hard. And suffering a drop in RPM in an engine designed to run flat out is a NO NO too. Hence the torque convertor wins out for drag racing.
==============
I think I am answering my own question. But I'm guessing a bit as I do it. Anybody care to shoot me down, use the right buzz words, explain it better, or whatever?
In top end drag racing, there is one hell of a lot of power, up to 7000 HP I believe in Top Fuel.
The main limitation is traction, not power, so they use a multiplate clutch tat is designed to slip to a mechanically controlled degree. This wastes the unusable portion of the power until speeds and down-force from the wing allow more power to be usefully applied.
The clutches are controlled by an arrangement of bob weights and springs.
More humble cars use a relative cheap and durable transmission. The transmission of choice is based on the old 1960's GM Powerglide. These use a very high stall speed torque converter, that allows the engine to be bought up to it's useful power band while the car is stationary on the start line, so the car launches with the engine in its power band, and the inertia to bring the engine up to speed is overcome before the race actually starts.
This transmission method is not nearly so efficient in terms of power in vs power out as say a 6 speed manual and a convectional clutch, but it more than compensates by keeping the engine in it's power band, and minimising inertia losses in the engine due to rpm changes during launch and gear changes. It also applies power without the interruptions of a manual gear change.
Other transmission methods used are manual boxes, but these only succeed in slower real everyday road car type classes, 3 and 4 and now even more speed automatics, but these mainly lack durability for anything more than daily drivers again.
Special race type manual type gear boxes are also used, with either centrifugal clutches, with solenoid or pneumatic application for clutch and gear box for gear changes, or with torque converters and air or solenoid shifters.
Some of the buzzwords are Crowerglide for the clutches, and Lenco for the gear boxes, and air shifter for the gear changer.
Basically Crowerglide type with a Lenco type and air shifter covers the top end, Heavily modified Powerglide covers the middle, and OEM whatever covers stock and bottom end classes.
Regards
pat pprimmer@acay.com.au
eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.
I should have used speed ratio, not slip ratio. My mistake for being misleading. A good clutch material will have nearly constant coeff of friction from 0.1 to 0.9 speed ratio, assuming constant apply force. Therefore the efficiency of the clutch is relative to the speed ratio in a nearly linear relationship. So, for example, at a 0.6 speed ratio, the clutch will be roughly 60% efficient, but a torque converter will be maybe 75-80% efficient.
In the big picture, you have to look at how much time is spent in a partial slip condition vs. time at 1.0 speed ratio vs. time at idle where a TC is worse (more idle torque) for fuel economy on a drive cycle. For performance, the TC has an inertia impact that may override its performance advantage. For drag racing, I imagine that a TC is easier to control at launch cause it is consistant run to run where a clutch will change over time (wear, mu) making it harder to control.
PatPrimmer, supposedly Peak HP has been measured with strain-gauge onboard TopFuel Dragster to be approx. 7933 HP
-------------------------------------------------------
30th edition of National Dragster, on page 56 there is an analysis of Doug Kalitta's 4.48 @ 333.9 mph. According to the data and calculations it peeked at 7933 hp on that run at 2.7 seconds into the run.
6205 f/lbs of torque.
------------------------------------------------------
my Program quick calculation;
it shows Top Fuel Dragster only hooking 2543.5 HP in the 1st Foot of distance...thats a lot of clutch slip ?
about 77.5% to 80.5% "average" total drivetrain efficiency over 1320'
RollOut Distance Launch HP = 2543.5
Engine Dyno Peak HP = 7935.8 {600 Rpm/Sec Accel Rate}
Engine Dyno Peak HP = 8221.5 { Steady State }
Air Density PerCent % = 97.11
Weather HP Correction Factor = 1.02980
DriveTrain + Rotational Inertia HP Loss = 1733.9
Net HorsePower = 3403.3 at 1320 Feet
at 334 MPH=
3355.9 Aerodynamic HP Loss
112.1 Tire Rolling Resistance HP Loss
897.0 Ram-Air HP gain
Best Traction 60 Ft = 0.8447 Max GForce = 4.332
RollOut ET = - 0.120 MPH= 11.383 GForce = 4.332
1320 Ft = 4.4800 MPH= 333.900
Maximum Tire Friction Coefficient = 4.332 GForce
Maximum Launch Lbs-Force = 9746.8
Note=> If RollOut Distance Launch HP is significantly
lower than Peak Dyno HP this usually
indicates a slipping clutch
==========================================================
from my previous Post on Jan 2,2004
the "total drivetrain efficiencies" were;
Auto Trans= 88.0 to 91.5 % Eff
Std Trans = 93.0 to 96.0 % Eff
where Differential Efficiency (including Driveshaft) is set at 97 % or .97
then..
Auto Trans =88.0 to 91.5 becomes .88/.97=.907 and .915/.97=.943
90.7 to 94.3 % for auto/converter alone, not including driveshaft and differential eff %
so for automatic/converter 90.7 to 94.3 % , Dyno to DragStrip correlation agrees very well with quote from
crysta1c1ear (Automotive)=>
"So let's say 95% average efficiency is round-about a theoretical maximum for planetary automatics in design, and since things never go as well as expected in practice, I'd say 94% average is maybe the best we will see, even if greater than 92%, ie 92% minimum (with caveats) is already being quoted by planetary automatic transmission designers."
for Standard/Clutch then 93.0 to 96.0 % Eff becomes
.93/.97=.959% to .96/.97=.99 % percent trans efficiency
95.9 to 99.0 % for manual trans alone
not including driveshaft and differential eff %
and some of ProStock Car onboard data shows about 99+% Clutch LockUp towards end of DragStrip
also in modeling DragRace Cars/Dragsters , my Program shows that auto/converter lightweight Dragsters with Hi-Stall have approx. 84.0 % total drivetrain eff (.84/.97=.87% eff)
..and heavier Cars around 3330 Lbs. have 90.0 % total drivetrain eff. (.90/.97=.928 percent eff)
again the 92.8 % agrees very well with crysta1c1ear (Automotive)
Larry Meaux (maxracesoftware@yahoo.com)
Meaux Racing Heads - MaxRace Software
ET_Analyst for DragRacers
Support Israel - Genesis 12:3
