How effcient are current automatic transmissions

[dadof275103]

I have just read that the Chrysler 904 Torque Flite is one of the most effcient automatic transmissions in use. Is this true?

How is this tested?

Are automatics still 4% to 6% less effcient than a manuel transmission?

Michael
dadof275103

 

[tbuelna]

Thecardoc,

CVT's are nowhere close to MT's with regards to mechanical efficiency. Although friction CVT's (like Audi's link-plate chain or VDT's push belt) are more efficient than traction CVT's (like Jatco's half toroidal or Torotrak's full toroidal), even the best friction CVT's can only get about 95% efficiency, input to output, when driving through the variator. This is mostly due to the parasitic losses caused by having to constantly drive the hydraulic pump for the variator pulley clamping and positioning.

Even an average MT will easily achieve 97% efficiency, input to output.

 

[Thecardoc]

Now granted I don't have percentages to throw out, but Audi's claims stand as THEY make them. The CVT out accelerates their manual transmission (and their automatic) , and achieves better fuel economy than both as well. Their claims, not mine.

 

[ivymike]

a car can get better fuel economy while using a less-efficient transmission, even if all other design parameters and operating conditions are held constant...

 

[Cargeke]

How does that work? If you're reducing the efficiency of your only variable, how can you increase overall efficiency?

 

[ivymike]

The efficiency of your engine is not constant, and the transmission you pick will influence the load/speed histogram seen by the engine for a given (whole vehicle) driving cycle. Use a less efficient transmission that puts the engine more frequently into an efficient operating regime, and you may have a net benefit in fuel economy.

 

[patprimmer]

To take Isaac's point to the ridiculous if you have a direct drive or 1:1 with no gears you obviously have a very efficient gear box, but when you need to operate in a variety of conditions and speed ranges, more gear ratios obviously help despite the drop in gearbox efficiency.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers for professional engineers

 

[Thecardoc]

Thanks for clearing some of that up. Now part of what Audi has told us is through the use of direct torque measurements as well as computed ones, plus monitoring input and output speed sensors they alter the pressure on the "chain" as required to improve efficiency. Under heavy loads more pressure can be applied to the chain through the pulley system to prevent slippage, under cruising conditions, this pressure is reduced, while still preventing slippage. The CVT even mimics a gearbox with pre-set ratios that can be commanded manually. (6 speeds)

 

[dicer]

CVT more effcient? How? No positive drive there, all slipage and wear. All that = loss

 

[Thecardoc]

http://www.sae.org/automag/techbriefs_01-00/03.htm

 

[NormPeterson]

Those Audi comparisons are still a little bit "apples vs oranges" in that no mention is made of any gear ratios, including final drive. Only a general note about gear spread inside the box. There is no way to tell how closely to optimum each transmission has been configured with respect to maximizing performance in the EU fuel consumption test (note that "optimum" was clearly in their quotes with respect to the manual transmission).

I'm not criticizing this development for what it is (even though I'm kind of hardcore about keeping my traditional manual transmissions).

The presentation for implying more than it should while providing less basis for independent comparison, on the other hand . . .


Norm

 

[Thecardoc]

Hi Norm.

Great points.

Would you believe? This material as outlined on that SAE page is the same information that is presented in the new product class for service technicians.

 

[NormPeterson]

I'm not at all surprised.

Enough knowledge to diagnose a component or system < enough knowledge to develop same. Particularly when mandated testing is involved and there might be some choice in what to optimize for.

Training for techs is only going to go to a certain depth even with the newer developments. While I do know a couple of techs who would understand this particular topic in much greater detail than that SAE page goes into (from briefly working at a dealership between engineering assignments), the majority probably would not, or have insufficient interest in even attempting to learn it (I met a few at the other end of the spectrum, too).


Norm

 

[tbuelna]

Thecardoc,

You are correct, the Audi Multitronic CVT variator does employ a mechanical device called a "torque sensor" to vary the pulley clamp load in response to torque. This allows the clamping force (and system pressure) produced by the hydraulic system to be lower, thus requiring less power. The torque sensor is basically a roller and helical ramp mechanism on the pulley shaft that creates an increasing axial force on the pulley sheave as torque tries to drive the rollers up the helical ramp. However, the torque sensor only reduces the parasitic loss of the hydraulic system. It doesn't eliminate it.

Consider this, if your CVT equipped car requires 30hp to cruise down the highway at 55 mph, and the hydraulic pump in your CVT is sucking up 1.5hp, that's a 5% loss. A MT under the same conditions would only have about 1.5% loss at most.

 

[izzmus]

In response to the original poster, I believe the "efficiency" they were talking about was the amount of energy it takes to accelerate the internals. So, for racing purposes, more engine power is available to accelerate the car. (Which is why you see people taking Chrysler's light-duty 3 speed and adapting it to, say, a Chevy big-block)

IIRC, the 904 also does something different internally compared with other 3-speeds, resulting in less lost energy during a shift. It's been a long time since I paid attention to drag racing in general or automatics in particular.

 

[maxaccel]

what do you think of this

http://www.torvec.com/products_ivt.html

 

[BrianPetersen]

Honda has a similar hydraulic-CVT as the above post that they developed for scooters and small motorcycles, and they have a 2009 model with it in production. I can't speak for the one above since it gives no details, but Honda's system combines a hydraulic swash-plate pump and motor in a single housing that rotates together with output shaft speed, and the whole deal is arranged so that at highway speed, the hydraulic part of it is locked up (1:1 ratio). At lower speeds when accelerating, the hydraulics transmit more of the load, and there is some inefficiency, but this is a smaller part of most drive cycles.

Torvec's claims of fuel efficiency improvements seem based on big reductions in engine speed, particularly at low vehicle speeds. The improvements are smaller as speed gets higher. It's possible to do that with practically any CVT and it's possible to do this with a normal transmission by upshifting really early, but most don't do it because of driveability concerns. (The 4 speed auto-box that they're comparing to, isn't a design that's going to set the world on fire.) You can have the economy, but it might not be worth the unpleasantness. I can get about 10% better economy than "normal" with my TDI by upshifting to keep the engine barely above idle, but it requires a feather touch on the accelerator to avoid lugging, and actual acceleration or hill climbing is nonexistent without a downshift.

 

[BrianPetersen]

can't edit: should note in above post that in my car (VW Jetta TDI) it's a 5-speed manual so no shenanigans required. The DSG version of the same car won't allow even the slightest engine load below 1500 rpm without downshifting.

 

[tbuelna]

izzmus,

At a certain operating condition, an AT with its coaxial input/output architecture, can actually be more efficient than a MT. With the torque converter locked-up and the driving planetary gearset locked-out, the AT simply becomes a "shaft". Passing the power thru without any speed change and with virtually no efficiency loss. The only parasitic losses coming from windage, clutch drag, bearing and seal friction and driving the pump. But they are minor losses.

A MT always must always pass power thru at least one gear mesh by design. So the MT would have slightly higher losses than a "locked-out" AT.

As to your comments about driveline polar inertias, the MT does likely have a lower system rotational MOI than the typical AT. But whether MT or AT, the transmission inertias are probably a relatively small component of the total drivetrain inertia. Especially when one accounts for wheels, tires, brake drums/rotors, etc.

 

[patprimmer]

MT at least used to have a 1:1 top gear where the main shaft was locked out and the lay shaft just spun with load except bearing and windage.

I thought 4 speed autos effectively had an OD top so third gear is 1:1 just as fourth is often 1:1 on a 5 speed MT

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers for professional engineers

 

[izzmus]

I cannot believe that any automatic can be more efficient than a rear drive transmission in Fourth gear (or whatever ends up being the 1:1). I believe this is why BMW used to use non overdrive 5 speed transmissions and unusually long-ratioed final drives, instead of a more normal final drive and an overdrive. (Of course, a transaxle must always transmit power through the geartrain since the output is not co-axial with the input)

This is all immaterial to drag racing, of course. Automatics have the advantage of not having to shift through Neutral to change gears, so you can never miss a shift, and the torque converter allows a much narrower powerband to be used with only two or three gears.

Incidentally, there are racing transmissions for circle track use that incorporate the clutch into the layshaft, and actually disengage the layshaft altogether when in top gear. No flywheel mass and no rotating gear mass or clutch mass when in top. These also usually have only two gears, one for use in the pits and yellow flags, and one for racing.