05 Legacy GT stuck in dry conditions

[whitetiger]

now, they do make upgrades to make this happen less(at least for the 5mt and 6mt anyways) but nearly no one here will go that far as they are mean for dirt rally conditions and will not be so great on the street(think axle bind).

[ShadowImg]

Adding a front LSD would certainly help quite a bit in that specific scenario. Of course, if you get both front wheels off, or one front and one rear, you're back in the shit.

 

I'm not 100% sure that VDC won't help in this scenario. It would sense the spin and brake that wheel I think, which would allow more power to transfer? I don't know how true this is, but it seems reasonable in theory.

[rob-2]

This is all very interesting.

 

I will attempt to duplicate this with a honda SH-AWD sedan later in sept and see if it suffers the same problem.

[rao]

Come one, we all know the Honda won't have any problems, it is not possible.

[bmx045]

Now you're asking for it.

[rob-2]

I am wondering if I had applied the brakes gently if that would have helped? Thoughts?

 

Could this have simulated tire contact?

[bmx045]

no, these aren't jeeps.

[outahere]

.............they do not lock up 100%. ......

 

According to the FSM, the center diff in the 5MT is capable of direct plate-to-plate contact "non-viscous operation" (i.e. lock up) when it goes into "hump mode". To get into hump mode will require that the silicone fluid be sufficiently heated, which means a few seconds of spinning tires will be needed.

[outahere]

actually, all stock Subaru trans' are vulnerable this way. from the lowly 4eat all the way up to the 6mt DCCD. these cars arent off roading rock crawlers with locking diffs. the wheels need to be kept on the ground.

 

In the linked video, a 2008 Legacy is shown (at around 8 minutes into the video) going up an inclined roller ramp with 3 tires on rollers (to simulate wet ice), and one tire on a solid surface. It shows the Legacy going up the ramp when there is traction on only one tire: first the LF tire, then the RF, then the LR. The tested Legacy has VTD and VDC. Only one tire needs to be on solid ground to keep the Legacy moving.

 

[ame=http://www.youtube.com/watch?v=t09ExAUgtyE]All ALL WHEEL DRIVES are NOT created equal. (Subaru Vs Other - YouTube[/ame]

[whitetiger]

According to the FSM, the center diff in the 5MT is capable of direct plate-to-plate contact "non-viscous operation" (i.e. lock up) when it goes into "hump mode". To get into hump mode will require that the silicone fluid be sufficiently heated, which means a few seconds of spinning tires will be needed.

 

 

even in hump mode, it can only transfer a percent of the power tot eh other end off the diff. the car will act as if the clutch is slipping badly. trust me on this. ive blown out my rear diff twice and snapped both my rear axles and still was able to drive it home. the center diff was indeed transferring power to the front end with the equivalent of the rear wheels being off the ground. But it was slipping like crazy and it would never have made it up a hill or over something that would have required more tq transfer.

 

go pull your drive shaft out and drive around to see what i mean. Hump mode is not a 100% mechanical lock up. Far from it.

[whitetiger]

In the linked video, a 2008 Legacy is shown (at around 8 minutes into the video) going up an inclined roller ramp with 3 tires on rollers (to simulate wet ice), and one tire on a solid surface. It shows the Legacy going up the ramp when there is traction on only one tire: first the LF tire, then the RF, then the LR. The tested Legacy has VTD and VDC. Only one tire needs to be on solid ground to keep the Legacy moving.

 

 

 

wheels on ice are not the same as wheels in the air. wheels on ice still give some traction and allow the diffs to work better. do that same exercise in the video but more inclined, and the car would not have moved.

[whitetiger]

I am wondering if I had applied the brakes gently if that would have helped? Thoughts?

 

Could this have simulated tire contact?

 

 

no, this only works if you have a torsen/helical active torque biasing diff. with those, the wheel with traction is given a biased tq amount compared to the wheel that is about to slip. the amount that is biased is based on the amount of traction the slipping wheel has. using the brake to simulate traction on a free wheel in the air will bias tq to the other wheel and move the car.

 

with these kinds of diffs, VDC works better since it will brake the spinning wheel and activate the diff to transfer power. this is why VDC in the 6mt sti makes the most sense. the STI has ATB diffs front and rear.

[iNVAR]

here is why it happened.

 

you assume that the center diff can transfer power to the rear if the front wheels are off the ground(even if one is off the ground, its the same as both being off since the stock front diff is open). the truth of the matter is the 5eat VTD center and the 5mt VLSD center are not strong enough to transfer full engine torque/power to one end if the other is open. its only gonna be able to do a fraction of it. and if that amount is not enough to move the car due to the situation the car is, then the car wont move. remember the bias ratios (50/50 for the 5mt or 36/65 for the 5eat) are just that, bias ratios. they do not lock up 100%. they only work like the bias indicates if all the wheels have some traction.

 

lets use the 5mt for example cause i know the numbers. the center VLSD in is is rated at 4kgf-m/100rpm(according to rallispec.com). this is the amount of force the viscous clutch can transmit from one end to the other in the case one end of the diff is totally open. notice it is based on rpm. this rating is independent of gearing or engine power output. it is only based on the difference in rpm from the front wheels to the back. so regardless of what your doing, if one of the front wheels if off the ground, the center diff will only be able to send 4kgf-m/100rpm to the back wheels to move the car. now that rpm is the prm of the output shaft of the transmission, not the engine, since the diff is on the output shaft. so, in first gear (with a ratio of 3.08 of engine rpm, just using the number as an example) even at 6000 engine rpm(you would never do that int he real world), the output shaft of the trans is spinning at about 1950rpm which means the diff will only transfer 78kgf-m to the rear wheels if the front had no traction. that may not be enough to move the car, and this is only if the center diff has no wear at all. odds are your diff is not as strong as it was new.

Something's not right here. I know 3.08 is just a number you pulled, but it's close enough to the 3.166 in the 5MT so I'll just go with your numbers.

 

According to this unit conversion:

http://locost7.info/converter.php

 

78kgf-m is 563.94 ft-lbs of torque... No way that's not enough to move the car. Something's gotta be off with the figures/calculations.

 

edit: Hmm, is it because I didn't reduce it by 4.111 (final drive)? That would make it 137ft-lbs at 6000RPM. Even at half the engine speed, you'd have about 60ft-lbs of torque. That should be enough to get the car MOVING at the very least.

[whitetiger]

the thing is, you cant just spin it to infinity and get to infinity TQ transfer. there is a limit to how much it can transfer not matter how fast you spin it. hump mode or not. i dont know what the limit is there, but im positive ther is one. ive driven with the rear diff essentially disconnected and it never locked up and drove the front wheels like ther was nothing wrong. it was like a severely slipping clutch and there was just enough power to slowly and gently accelerate the car.

[iNVAR]

Yeah, I understand, but even you're saying you were able to gently accelerate. From what the OP is describing, he should've been able to get moving, albeit slowly. It doesn't sound like he was pushing against any major type of resistance, certainly not any that should stop him from slowly moving (in theory.)

[dmanaenk]

edit: Hmm, is it because I didn't reduce it by 4.111 (final drive)? That would make it 137ft-lbs at 6000RPM. Even at half the engine speed, you'd have about 60ft-lbs of torque. That should be enough to get the car MOVING at the very least.

You should multiply those 563.94 ft-lbs by 4.111, not divide. Then divide by wheel radius.

To get the car moving you gotta know the incline he was on. He claims flat ground, so very little torque at the wheels should've been enough to get moving. One should be able to manually move the car when it's in neutral. I'd say 50kgf/100lbs should do it to overcome rolling resistance. So, say 100 ft-lbs torque at the wheels, or barely 25ft-lbs at the driveshaft.

 

Old 4EATs had an FWD fuse, that some sneaky used-car dealers used to hide 'torque bind' problems on higher-mileage cars. Seems like smth similar here. Maybe some solenoid is gone or smth.

Could also be TCU glitch, since it's 'predicting' what the car's gonna do an deciding on the torque split.

[dmanaenk]

the thing is, you cant just spin it to infinity and get to infinity TQ transfer. there is a limit to how much it can transfer not matter how fast you spin it. hump mode or not. i dont know what the limit is there, but im positive ther is one. ive driven with the rear diff essentially disconnected and it never locked up and drove the front wheels like ther was nothing wrong. it was like a severely slipping clutch and there was just enough power to slowly and gently accelerate the car.

I wonder at what temp your center diff was. If only a fraction of engine power was transmitted to the wheels - the rest shoud've gone somewhere. Heat, I guess.

[iNVAR]

You should multiply those 563.94 ft-lbs by 4.111, not divide. Then divide by wheel radius.

To get the car moving you gotta know the incline he was on. He claims flat ground, so very little torque at the wheels should've been enough to get moving. One should be able to manually move the car when it's in neutral. I'd say 50kgf/100lbs should do it to overcome rolling resistance. So, say 100 ft-lbs torque at the wheels, or barely 25ft-lbs at the driveshaft.

Ah. My knowledge of physics is very rusty and limited, but now I see.... Torque multiplication since it's a 4.111:1 reduction in rotations. Thank you.

[whitetiger]

I wonder at what temp your center diff was. If only a fraction of engine power was transmitted to the wheels - the rest shoud've gone somewhere. Heat, I guess.

 

 

probably very hot.

 

ive had either the rear diff go or the rear axles snap and ive made it home all 3 times, though each time it was harder, the viscous unit is probably pretty worn out by now. replacing it has been on my list for some time.

[crossle32]

Next time, try riding the brake, gently- it is harder to do in a 5mt but easy in a auto. This should slightly keep the free wheel from spinning, there-by transferring some more torque to the wheel with traction.

This method also works in an open diff rear wheel drive, in the snow. Slightly pull the hand brake to stop the spinning wheel and transfer torque to the wheel with traction. This only works until both wheels start spinning.

 

Contrary to what others say, this does work. Not always, but has gotten me out of a few slick spots. No theory, practical application. Also, if I recall VDC did not come out until 2008??

[whitetiger]

^that only works with ATB/torsen/helical diffs not open diffs. with an open diff, even if the wheel in the air has some resistance, the open diff will still send all the power to the wheel in the air since its still the easier wheel to spin.

[frank_ster]

an 05 lgt will spin 1 rear wheel and one front wheel. if neither of those have traction the car isn't gonna move.

 

the newer modles with traction control will apply the brakes to the spinning wheels.

 

yes i know it sounds dumb. but you can actually get a lot further than you could of with a front wheel drive car.

[CTATV]

What year car do you have ? My 05 Gt Wagon 5mt doesn't have the button. But my 09 Spec does.

 

OP put more info in your profile

 

07 first year for I-drive and for traction control.

[dmanaenk]

an 05 lgt will spin 1 rear wheel and one front wheel. if neither of those have traction the car isn't gonna move.

Would you care to explain the mechanics? Say, for a 5mt 05 lgt.

[__raj]

My observation of having two wheels on the ice and the other two on dry pavement is the iced ones spin on my former 04 WRX 5mt and current LGT. Occasionally power is kicked to the dry ones and the car moves a bit. However always one rear and one front spins.

 

I am curious how well our new to us 07 Acura MDX works with SH-AWD. Apparently it kicks power not only front to back but can go to a single wheel(rear) or one side full side. I noticed in hard turns when really pushing it actually puts power to a single rear wheel with the AWD display.