I think you’re really hung up on the “it’s not a differential!” thing. A differential isn’t the only thing that can distribute power from one source to two outputs in varying amounts. The electronic clutch system isn’t a differential, but it serves the same role as one. Seriously, read almost anything about the electronic clutch system and it’ll tell you all about it shunting power back and forth as needed, not just switching on and off reactively.
The STi has both working in concert because the mechanical differential is more responsive around corners and such but the clutch system can be used to control the torque split electronically.
Same thing as always. Longitudinal layouts are better for normal-50/50(ish) AWD than transverse ones.
Where exactly? If it’s because as you claim only the manual transmission ones have a true 50/50 power split, why have those also made big gains in fuel economy?
Err, oops. But where are you seeing the description of the earlier Sienna’s system? Everything I see implies it’s the same as the same vintage Rav-4, which is a normal 100% FWD, then up to 50/50 re-actively.
So the Cobalt’s manual is an extra 2 MPG. The same year Impreza is only 1 MPG. Big whoop. The same year Corolla is exactly the same for the 4AT and 5 speed manual. The same year Civic with (granted a 5 speed) automatic beats the manual by 2 MPG. I’m not being disingenuous here, I’m just not seeing the broad trend you are.
My understanding is that a traditional 4WD transfer case is effectively not symmetric. My '79 Subaru Brat info said that the rear axle ratio was not the same as (lower than) the front. Hence, if I pulled the 4WD clutch lever to send power to the rear wheels, doing so on dry pavement, even in a straight line was a bad idea, because the front and rear wheels would be scrubbing toward each other – I tried it once, and it did not feel pleasant. One mechanic informed (or misinformed) me that almost all part-time 4WD systems have scrub, either inward or outward, and so are not suitable for dry, paved roads in 4WD.
Your Subaru Brat actually had part time AWD, they just did not mark it as such because a large number of consumers would never have understood what that meant.
Hmm? The Brat was pretty much a straightforward part-time 4wd with fixed ratios. Although really at that point the AWD vs. 4WD distinction hadn’t really developed. For example, the AMC Eagle used a full time system that would probably be called AWD today, but was never badged as anything other than “4 wheel drive.” Even as Subaru switched to the full-time system on most of their cars in the late 80’s, they still called it 4WD for a while, although their badges were always kind of 4-that-looks-like-an-A-WD.
Since this is coming from someone who doesn’t really seem to grasp how a limited slip diff works (hint: except in a torsen, there’s always some clutch-like mechanism, the viscous fluid in mine acts like a clutch and wears like one), I’ll take this as a compliment.
And if by some miracle you actually understand how one works, and are just explaining yourself poorly: how does the center split in an auto Subaru work if it doesn’t supply a “differential”?
The fact that it had “ratios” indicates that it was AWD, part time 4wd has transfer case which is a locked spool when it is engaged, there is no torque ratio as it will transfer torque at any ratio just as long as the two driveshafts turn at the same speed. I will do so up to the point where something physically breaks.
The earlier Subaru clutch pack and later gear based fluid coupling center diffs are similar to several “all time 4wd” at the time. Using the most common current usage of 4wd describing part time 4wd with a locked center diff it would be a part time AWD IMHO.
The “ratios” we’re not in the transfer case. All the transfer case has was a power take off shaft that clutched to the rear drive shaft. The ratios I spoke of were the difference between the final drive ratios of the front and rear diffs.
There is typically no difference between the two, the need for scrubbing of tires is due to the tires traveling different distances. The gear teeth in the front and rear diffs need physically large teeth to support the driving torque and thus do not have enough teeth to do any type of small gear difference.
E.G. 4.11s have 37 ring gear teeth and 9 pinion teeth 3.73 have 41 / 11. Plus there is no technical reason to have gearing differences in the front and back. When Subaru says 60/40 they are talking about how the limited slip center diff is biased compared to an open diff. An open diff will always have a 50:50 bias but that is torque and not “work” or rpm.
Bear in mind that the front diff was integrated into the transmission housing. The transmission output shaft had a drive gear on the forward end that engaged the full-time front differential. The tail of the output shaft fed out to the transfer clutch that could manually engage the driveshaft to the rear wheels (some articles describe engaging 4WD as “push of a button”, but mine involved lifting a lever). So having a stock rear differential that was not exactly the same as the integrated front differential would not necessarily be surprising. I know what my Haynes manual told me, and I know how 4WD felt on dry pavement.
Do you have a cite for this? As far as I know, the Brat (along with all early-to-mid 80’s 4WD Subarus, barring some exotic stuff) was just a regular part-time 4WD with a fixed front-to-rear ratio. (And, yeah, I guess “ratios” was a typo.)
So yes, a clutch can be a clutch, you don’t need a differential to sometimes drive the rear wheels (or front wheels), you don’t even need a clutch either, a pickup truck’s 4 low doesn’t have a clutch.
However a clutch and a differential are still not the same thing. Greasyjack, the link you posted just says that AWD and RWD have more drivetrain losses than FWD, by “up to 50%” in fact, if you compare straight FWD to full time AWD, sure, but that’s not your claim. Your claim is that:
A longitudinal layout has all of the losses of the RWD drivetrain, plus it also has a 180* turn and then another 90* turn to power the front wheels. How is this better than the transverse drivetrain, regardless of whether 2 or 4 sets of wheels are being driven? The claim that I had trouble with was that a spiral type gear set that rotated power to a different axis was significantly more lossy than a gearset where both gears rotate on the same axis. The article doesn’t say this, only that “losses of x percent are the norm”, well is that higher or lower than gears on the same axis, and if so by how much? 1%? 10%? Enough to make one layout markedly superior to another?
And just so I am clear on all your positions here, do you agree that a differential and a clutch are not actually the same thing and in fact a differential can do things that a clutch cannot, hence the reason why some cars (like Subarus!) use one or the other depending on the vehicle? That’s really what I’m saying. You seem to be on the one hand arguing that a clutch and a diff are at least functionally the same, or are equivalent in capabilities for the common usage scenario on non-race cars, which is fair enough, and yet on the other hand, arguing for the superiority of the Subaru system, when in fact the automatic Subaru system works in exactly the same way as most of the other clutch based systems on the market. A clutch system by definition must be reactive, unless your clutch has eyes and can scan the road ahead of your car. It doesn’t need to only react to slip, which is the claim that many people make of them, they generally also react to things like throttle position and steering angle, but again none of this is unique to the particular clutch system that Subaru uses. Floor the throttle in any AWD vehicle with a clutch based AWD system and likely the clutches will lock together.
There’s plenty of planetary gears in that transmission to move the torque around. Even if it’s not in the arrangement that we normally recognize as being the contents of differential pumpkin, it performs the same function using a set of gears. The slippage necessary to divide power is provided via the the clutches. Without them (or something functioning as them), it’d be a locker, like 4 low normally is, instead of a limited slip.
So, if you expect it to be a classic pumpkin diff, no it’s not. It’s a differential in action, though.
That was simply in response to your request for a cite that gears that transmit power at an angle have more losses than one that transmits them on a plane.
But here’s how that shows my greater point: as per that link, drivetrain losses are 50% lower for FWD because it doesn’t have to make the 90 degree turn. But on a transverse AWD car, the power going to the rear has to make two 90 degree turns. So you get three times the drivetrain losses sending power to rear than you do the front. Power going to the front and rear on a longitudinal layout both have to make one 90 degree turn each, so minus the losses in the power-split device itself and such, it’s basically the same as a RWD car. And, importantly, unlike the transverse layout there isn’t a huge efficiency penalty to sending more power to the rear. From a driveline losses standpoint, it’d be the same no matter how much power it sends to each axle.
Two gears on the same plane turning in same direction is pretty much simplest way to transmit power with gears imaginable. I really don’t understand why it’s so difficult for you to believe that they’re more efficient than gears that have to transmit power at an angle.
(My bold)
Firstly to address the bold part. What the heck does this mean? By that logic a center differential is just as reactive. Once again, what is usually meant by a “reactive” AWD system is one that is entirely or almost entirely FWD until slip is detected, at which point it sends some power to the rear wheels until the wheelslip resolves itself. This describes a large portion of transverse FWD-based AWD system. It does not describe anything currently sold by Subaru.
The system on automatic Subarus is, yes, also reactive if you want to be pedantic but the difference is that it’s normally close to 50/50, and “reacts” by changing the torque distribution to allow the car to go around corners. In terms of how useful an AWD system is for improving handling on and off road, this is a HUGE difference versus one that just occasionally kicks some power to the back**. The Subaru system most definitely does not “work in exactly the same way as most of the other clutch based systems on the market”
To address the above quoted part more generally:
Yes, obviously a differential as in “a bunch of gears that take power from one output and mechanically transmits it to two outputs in varying proportions” is different than a clutch as in “a piece of friction material that can engage or disengage the transmission of power.”
However, we’re not talking about a single clutch, we’re talking about the whole kit-n-kaboodle that is the multi-clutch system, which uses multiple clutches and hydraulic controls to do, yes, essentially the same thing as a differential. People sometimes even do call them electronic center differentials, which is technically wrong, but it wouldn’t be wrong to call it a pseudo-differential or something. Now, yes, “real” mechanical differentials and these electronic pseudo-differentials do certain things better and worse, which is why some cars have one and others have the other and some expensive cars have both. But broadly speaking they both: take power from a single source and distribute it to two axles, with the ability to vary the amount of power dynamically.
** Also, I’ve come up with a retort to your “anyone who owns or drives a Subaru is almost certainly less knowledgeable about AWD than the average population” jab: Yes, you learn a lot more about how AWD works driving a shitty one than one that actually works well. I live somewhere that it snows about 80 inches of snow per year. I’ve driven plenty of shitty AWD cars and you are keenly aware how the system works if it keeps breaking and regaining traction as the reactive system cycles, or if you’re trying to climb a slick hill and a front wheel is spinning but the system isn’t sending enough power to the rear to push the car up the hill. I would have missed out on many such educational moments if I’d been driving a Subaru from the start!
A mechanical differential not reactive, an axle with a center differential is always sending 1/2 of the torque to each wheel, full time, as it were. One wheel can even turn faster than the other.
A clutch or any other coupling device can at best only make one end spin at the same rate as the other. Even if you kept the clutches 100% locked up when going straight, you still have to decouple them at least somewhat for the car to go around turns. When this happens, more torque will go to the front (let’s stick with a FWD biased system like Subarus as an example) wheels than the rear. What happens when torque overcomes traction, especially when the wheels are also using some of that traction to turn the car? Wheel slip.
With a mechanical differential, the torque is always split equally, therefore all things being equal, the torque at each end will be correspondingly less than if they were all concentrated in one end, less torque, less chance of wheel slip.
A very old, wise and experienced off-roader* once told me you never mount the winch on the front of a truck, it will only pull you in deeper.
*Seriously, this guy won his class at the first Baja race in a Rambler, (he knew Baja better than anyone else racing that year), and he has spent more time in low range than 90% of off roaders have spend off pavement. He has also authored books on the subject. Guy is hard core.
Yes, the car can’t go around the corner if it’s 100% locked up. So the computer doesn’t lock it up 100% when it’s going around a corner. The difference is that normal reactive AWD systems are 0% locked up going straight down the road whereas the Subaru one is something like 75% locked up normally. The clutch packs are designed to slip continuously, and there’s always going to be enough slip in them to accommodate the difference in axle rotation speeds. It has the ability to send more power to the rear by increasing the clutch pressure and more to the front by decreasing it.
Again, no, it isn’t a differential. It can’t send full power to the rear axle like a differential, or even 50/50 through a corner. But it can provide full-time power to both axles at a ratio that’s relatively close to 50/50, which for most drivers is functionally the same thing.
Oh, also I actually thought of a somewhat frugal car that came with a balanced transverse AWD system! The late-80’s early-90’s Corolla All-trac wagons had a viscous coupling center differential similar to the Celica GT’s. 28 MPG for the FWD version, 22 MPG for the AWD one. Same motor, same gearbox, but 6 less MPG! The same year Subaru Wagon (with a stick, so viscous coupling) was 24 MPG for the FWD, 23 for the AWD. The same year Audi 100 was longitudinal and got 19 MPG FWD and 18 MPG with the (50/50 split torsen) Quattro. As I said, the penalty for default balanced (or near balanced) AWD is much steeper on a transverse engined car than a longitudinal one.
A winch should really be used to keep yourself from getting into trouble, or to pull someone else out of trouble (e.g. get someone started out of a sand bog). Relying on a winch to pull you out of trouble is like jumping out of aircraft with only the reserve canopy packed. In that context, a front mounted winch, where the operator can have eyes on the object being winched, makes more sense. Also, most older and heavy duty winches use an engine power takeoff (PTO) or capstan winch so mounting it near the engine is easier although I have seen rear mounts, and even in one case an elaborate center-mount on a BJ45 (winch was mounted to the frame rails through the deck of the vehicle with a feed through below the passenger seat and around the engine to the front, or straight out the back. The winch in this case was generally used for safety on river crossings, so both fore and aft winching (to pull across other vehicles) was desirable.
I have to ask because it really isn’t clear to me; shifting torque between the front and rear has absolutely nothing to do with going around corners, per se so what is the argument here?
If an AWD system has a centre diff (and if that centre diff is a viscous coupled version it almost acts like a clutch, varying the viscosity between the plates to balance torque), it will transfer as needed when the front and rear diffs are moving at different speeds.
The cornering issue is the difference in speed between the inner and outer wheels, which is addressed by either having an open differential (usually geared) or having some sort of limited slip system.
To my understanding of it, the main difference between AWD and 4WD is a matter of what is now called part time systems where the center, rear and/or front diffs can be locked to provide an even split of torque on all 4 wheels and the “slip” is at the tires interface with the ground (which is why my Cherokee didn’t like going around corners in 4 low on dry pavement) vice a “full-time” system that introduces that slip in the diffs themselves via viscous coupling or clutches.